Misplaced Pages

On the whole, arbitration works well. So I wouldn't want to change it wholesale. But there are several edges that it's probably common consensus can be improved and would be widely appreciated to improve.

• First - A first priority has to be "don't leave it worse". That means, for example, in the haste to fix a perceived problem, don't leave it actually with more serious problems for 2009 to fix. Don't let what does work well at December 2007, slip -- and in many ways, a lot does work well.

• Speed and responsiveness - Then, look at it from users viewpoint. I see a few key areas where some kind of change could be possible, and in which many wish change would happen. Speed and responsiveness is a big one. I work in business, and I'm used to starting in january with a team that feels a 30 day turnaround is about fine, and by June we're working to a turnaround under half of that. Likewise, we've got used to a 6-8 week turnaround on some cases. There's no reason for more than 4, and in many cases we should be able to break 2 or 3 if we discuss and agree how to case-manage them, the timings and processes to establish, and so on. Probably there needs to be discussion within arbcom (who know what the issues are for real) and also communal discussion of the suggestions (to get consensus that allow improvements). I've discussed this with several of the other candidates already, it seems that thoughts of these kind will be held by more than one person. How it works out is down to human nature and interia, but in either case, 1/ I'll be using the skills I have to try and do so, and 2/ if it can't be done I'll be fine summarizing factually, an overview on why not, because the community will deserve to know why a body it has delegated to, isn't delivering as people believe possible -- whether to fix it, collaborate to agree a better way, or at least understand why it wasn't practical. Arbcom is answerable to the community, not the other way round.

• Transparency and accessibility (= communication and contact) - can probably be improved. Or at least, the perceptions of these can. Arbcom and its members work phenomenally hard. The cases, and the additional invisible problems by email, are a very heavy workload, enough to burn out many experienced committee appointees in a few months. A lot of what Arbcom does must be and remain, a "black box". But this does not mean "set aside communication and explaining". In a way, communication is one of the most important things one can do. Good communication seriously quells disputes, heals divisions, assures all parties they are heard and respected, prevents people feeling that arbcom is "remote" and un-clued-in, explains how good decison making works and what policies mean, and a dozen other things. As an administrator when I started working on AFD, I began the habit of writing full length closes. They were wordy, they were long... but they worked. Over and over, people in the most contentious disputes commented afterwards that they felt it had helped, made sense, clarified what would otherwise have been assumed to be a biased decision by some side or other that it really was carefully considered and so on. It worked in the toughest disputes, almost as routine, it was appreciated by those declined as much as those supported, it took people who were flaming and helped them get to "I understand now", even if they didn't always like it. So I think Arbcom can communicate better, just because I know I do so and have done so, and I'm seeing what it achieves. Not unduely, or to breach privacy, but simply to bridge the gap non-arbcom editors sometimes have felt and described (despite arbcom's best efforts), of communication, contact, and trust.

• Decisions in tough cases - I'm told in a few cases, the committee, like the community, was split and indecisive. That's to be expected, but a direction, principles and ultimately a decision must be found in such cases, and this must be a decision that gains wide respect from the community. Again, its down to insight, draftsmanship and focus, and that's an area where I have a significant advantage - I've done the drafting for an unusually large number of policy pages, disputes, and sensitive complex matters, a fair bit of my offline life is spent figuring out "whats really going on/what might help this work" in problems, and I think I can help a lot there by simply doing the "donkey work" such as drafting and case summarizing, which I do already on tough matters. I find a good case summary and analysis makes a decision almost obvious, in many cases. If you want a few examples, let me know.

• Conduct while a case is being heard - in cases of habitual bad conduct and edit warring, problem users often get an effective 'free hand' from Arbcom to persist, while the case is considered and discussed. Its good evidence of conduct, but a case just doesn't (or shouldn't) be accepted at arbcom without good evidence there's a problem already. So there's a limit how much it's helpful to allow more. Especially, people come to arbcom specifically to have problem behavior dealt with. I'd not mind seeing Arbcom issue brief warnings and then temporary injunctions, in cases where there's ongoing escalation or continuing misconduct by an involved party during the case. People can die waiting for the ambulance, so to speak. Or in Wiki terms, become disheartened waiting whilst the bad conduct continues. We can do better, I think, by ensuring once a case is accepted, the worst ongoing abuses (if any) will be addressed even as the case is heard, and not (apparently) ignored.

• Process and administration - There's a question if the committee is operating the best way - right number of people, right length of appointment, right handling of absence, and so on. I've participated in that discussion so far, and will in future. The questions are things like - does arbcom have the capacity for its present workload? Does it have enough spare capacity not to burn all its members out? Should there be more non-arb checkusers? Should arbcom run two or more "circuits" or should it have one as at present? Should stay on arbcom reduce to a year or two, can members take a break, and what process is used to replace people temporarily or permanently? If someone is absent, how long do we let a case be held up? Is it taking on tasks it shouldn't, or not handling any it should? Things like that. I'm as much on the arb mailing list as you; so I would have my own understanding to gain from those with experience before forming too many fixed opinions. Again, this is what I do for a living, so I'm comfortable looking for ways to help it be better.

• Standing of committee - Last, one other for now (there's more but these seem to be some of the main ones). The last one is, does arbcom have a role just as a jury/judge in disputes and an advisor in sensitive matters? Or does it also have a de facto leadership role of setting certain directions for the community? In what areas should arbcom have more "say", or less, and who can override who? That sort of thing. I think arbcom's done very well here, but the relationship between committee and community is always going to be one to watch. What does the community expect arbcom to be?

I would never say that was "all". Like the community and project, Arbcom grows, matures and changes. It may have internal processes none of us see that prevent it doing its job at times, for example. It has a range of people who must work together. So the above is said with certainty that if I sat down 3 more minutes I'd find another few ideas, and knowing that many people's top wishes aren't on that list.

(There's one to do with sensitive information handling which I know is going to be discussed, but I don't know if it's my place to more than allude to it right now, for example. Another is whether editors who through no fault of their own have been attacked or smeared unfairly in cases that come to arbcom, are sufficiently vindicated by arbitration, or whether more needs doing from this direction. A third is in cases themselves, whether fact findings and remedies are always well chosen.)

But if the January committee do no harm, leave it no worse, and do address some of these kinds of things effectively during their time, then we'll have been good custodians the coming year and have improved.

I've answered a fair bit of the "what do I think about arbcom (including case duration) and what would I do" above (1). Take a look, and if you want me to expand on any part of that comment, do let me know, I'll be glad to :)

Do Arbcom act too hastily, and facets get dropped...? I'd like to say that on the whole, arbcom seem remarkably balanced at what they do, especially given the complexity of the cases they look at. However this isn't a single-sided question; the question covers several points:

1. Do arbcom get it right, but fail to communicate this, so that people perceive an error of judgement?
2. Are some errors in fact legitimate differences of interpretation, or application of good faith?
3. Do arbcom take the decisional view that rulings aren't expected to be perfect (given the pressures and complexity) but can be reviewed if needed, whilst users expect perfect insight?
4. Are case pages so messy by the time all evidence and counter-accusations are presented, that it's not clear where a decision has come from or if it's well informed?
5. Do arbcom have their own views and criteria to measure what constitutes "successful/appropriate case handling", so that the targets they work to are slightly different to the broader community's? (Ie, with more experience comes hopefully, wisdom in deciding what a "successful" handling might be, what matters, and what doesn't. Conceivably facets might well be dropped through wisdom and clear thinking, and not error, in some cases.)
6. In some cases, is there non-public evidence involved, so that what seems unusual from outside is a partial view? (By its nature, decisions based on such evidence may not be capable of informed review by other editors; this is a communal status quo with high buy-in though.)
7. Are cases well-assessed for acceptance or rejection? (It's easy to look at cases once accepted; what about the acceptance decision though?)
8. Last, the genuine concern: are there cases where significant aspects or points of principle are missed, misunderstood, or mis-assessed by arbcom, or poor judgement arises in the proposals and decisions, that are not covered by any of the above?

The striking thing is that 6 of these 8 essentially speak to communication and communications failure, rather than case errors. Most of these can easily be stated, discussed, and reassured, rather than left to create silences and perceptions of concern to others. So quick answers to the above:

1. Perception and communication

In my own experience, there are regularly cases where the involved parties don't seem to be reassured that arbcom did a good job (even if they did). Cases come to arbitration after a long history of stress, and good faith parties usually want above all, reassurance of good handling and a good final resolution. So if we can reduce stress and reassure people better, then we should. This is a matter of communication as much as decision-making, and I think arbcom could do a lot better at communication. Reassurance and certainty aren't trivial if you're in an arbcom case, and if users did have a serious concern something's been overlooked, we can certainly explain why it was handled that way, or look into it further. There are routes for such communication, but they're not adequately used.

Personal example - communication    Taken from the 'Attachment Therapy' arbitration case (DPeterson and socks, July - Aug 2007)

One experience of this I saw close-up this year was in the DPeterson and socks case (July 2007) where the sock-warrior was allowed to run rampant with untruthful smears and personal attack and misrepresentation, and yet no proposals were being added to clarify that arbcom realized this. In fact Arbcom almost didn't communicate at all, in the case. It would not do harm to be more communicative. As a result the good-faith parties I was helping (and myself) were in fact left very concerned that they actually didn't understand the complex evidence, or were going to be bamboozled by this editor or handle him hopelessly inadequately. A 1 year ban proposal also gave the editors new to arbcom much concern, because from where they stood, they saw only that this was a dedicated warrior who wasn't going to change, and interpreted the result to mean Arbcom grossly underestimated his virulence. What was missing was an explanation that there had been understanding, but that a years ban would be rather tougher than it seemed, and the community would handle the rest. Had I not explained, the parties concerned would have taken this impression of Arbcom with them. So a lot of the concern over "dropped facets" in some cases are actually avoidable, and based on poor cross-communication. I made a start on this by writing Arbitration policy/Case handling which explained these kinds of things for newcomers to Arbitration. The rationale was, as a non-arbcom member I couldn't make arbcom communicate better; but I could aim to ensure people visiting arbcom knew what to expect and why, and therefore be less worried by some features of current case handling. If appointed, communication during the case (a willingness to consider questions regarding handling, progress, and where it's at, for example) is something I'm fairly sure I can improve, and would remove much stress from a lot of cases.

2. Are cases wrongly accepted or rejected

Yes, I think so. I'm going to put myself on the line a bit, noting at least two cases I felt were accepted wrongly. In each there was either no case, or no case needing Arbcom attention. With a clear statement, this could have been explained. One was the BJAODN "Wheel war" case (later dropped following my suggestion ), and the other was Sadi Carnot. I feel the Sadi Carnot case could easily have been handled at ANI and probably got rapid consensus, if Arbcom had explained why they felt this. The problem in this case was that people jumped hard, and it escalated quickly, and once there was a block, unblock and differing views, it was sent directly to Arbcom as "divisive". In fact within a fairly short time it became obvious that there hadn't been wheel warring (by consensus definition), there weren't major problems, the users spamming was often not his own actions, and the user had shown willingness to engage in dispute resolution but had never been asked to. This was actually a pretty straightforward COI/POV pusher who had been mismanaged and then suddenly became a major headline. I'm not very happy at the ruling or lack of prior discussion either, but that's a different matter; and note it may easily have eventually come to a ban or long term block by other routes too. I wrote a full analysis of the background of the case, and evidence to back this, at the point it was looking to be accepted by arbcom, and I still feel this case (if returned with explanation to ANI) would have been adequately handled there, at that point. There was no great difficulty or complexity to suggest that Arbcom was really needed.

That said, it's better to accept a case in error than reject one in error.

3. Do facets actually get overlooked (I)

It's almost inevitable that some will be. These are complex cases and the evidence is often very long and disputed. Arbcom aren't lawyers or legal draftspeople, so rulings and such do occasionally have gaps. Evidence of this can be found in cases that return for clarification, or "fixing" of rulings. I think arbcom would argue that in many cases, the ruling isn't expected to be perfect; rather it's expected to give enough that if the problem continues it can always be reopened for further review. That's speculative but would make sense. Obviously one prefers to get it 100% at once, but cases are complex enough that I can see "do what we can readily do, and accept if its still a problem later" having been deemed a plausible strategy in some cases. I'm not sure I'm entirely comfortable with less than high quality decisions, but even on Arbcom one would expect different approaches. Again, communication would help; it would help if people were aware, so there wasn't this concern over it. There is no real reason not to be more forthcoming about the basis on which cases have been assessed and allegations considered, it need not be drama-inspiring, and tends to unite rather than divide.

4. Do facets actually get overlooked (II)

Does arbcom ever fail to take cognisance of key evidence that they should notice, or misjudge the case and users involved? I'd be amazed if they didn't, at times. The canonical example here would be a warrior who is skillful enough that they can pull the wool over arbcom, or the evidence is confusing and poorly presented, or where frustration by the other party has led to both being less than perfect actors so both get sanctioned rather than the genuine problem user being seen for what they are.

Of the two kinds of error - penalizing someone who did little wrong, or letting someone off lighter who did - the Committee seem to err on the right side. Rarely do arbitration cases seem to make the error of penalizing someone wrongly. Usually if there has been error, it's under-estimation or lack of appreciation of seriousness, or telling both to improve and not end up back at Arbcom again, ie, the fixable kind.

So for example, there have been cases where a clearer statement was wanted and instead Arbcom returned the community a ruling that put them back where they started without any further help. There have also been cases where all parties are told to start from fresh, and this seemed to some like an error of judgement. There have certainly been rulings that (in retrospect) were inadequately drafted and needed fixing. and cases (like the Albania-Azabaijan ones) where some serious sorting out was needed that didn't happen. What is notable is, that all of these tend to work out well eventually (or can be modified later if insufficient), since what was returned to the community was enough to provide foundations to help somewhat, anyhow. At least they do not often seem to produce manifest miscarriages of natural justice.

I don't think "hastily" is a word most would use on arbcom cases, and it does tend to have a significant payoff: - cases where the actual case has miscarried significantly in the end, to unfair detriment of an editor, are in fact uncommon.

(That said, see (3) below for two examples of genuine case mishandling. However, one of those cases didn't misjudge evidence, it just refused to rule, which is erring on the side of caution, and the other is questionable as to the harm done - maybe the user would have been banned, maybe not, with better Arbcom handling.)

Hope that helps! FT2 05:02, 22 November 2007 (UTC)

Two recent cases come to mind:

Allegations of apartheid

The findings in Allegations of apartheid were obviously a source of problem to the Committee. Several sets of proposals were presented by Arbitrators seeking to pin down the best proposed wordings, but although arbitrators were able to identify what was wrong with various wordings, they found it exceptionally hard to propose a wording that was right. One need only to look at the "opposes" on the proposed decision page to see the issues identified in different approaches.

For example, when I reviewed the proposals, it was clear the proposed principle on IAR (2) was not in fact anything like a good representation of the "limits of IAR", and finding of facts 2 , 3 , 4, and 6, all floundered because, although the findings centered around using article creation and AFD to make a point, the committee hadn't thought to lay the groundwork for this by including a principle on what the AFD process was, and how it was to be used.

As a result several proposals -- including all remedies -- floundered after that, because they couldn't find a way to clarify the concern neutrally from all sides. A proposal for a general amnesty also floundered, and rightly so, because it was clear that significant untoward conduct had taken place that needed specific comment. I tried to add some help by drafting a better set of proposals for the problematic area, but in the end Arbcom just didn't seem to find a way to address the case.

The end result was the case was effectively dismissed because Arbcom couldn't work out what to say... which is so far from ideal that I'm almost at a loss. For example, the case lingered for almost three months (11 weeks) in committee, with not one principle or finding of productive output as a result. The matter was ultimately deemed closable as "being resolved by the community", but this does not equate to any "success" for the arbitration process. Some general findings and measures might have been suggested for this significant dispute. Arbcom is there precisely to help guide a resolution to such matters, and the general principles that led to this case arising were not, in my view, as well addressed as they could have been. I'm not aware of any novel or off-wiki issues that were relevant. It was not an exceptionally complex case to need such indecision.

Sadi Carnot

This case was an editor who was, clearly, a link pusher, with COI, SPAM, POV/"own theory" issues, and using citations in a dubious way to "support" these, going back to 2005 or earlier and a previous account.

Unquestionably there were genuine issues and concerns. The problem was that the case was "railroaded" (I think the term is) - ie, it wasn't a concern, no formal or informal talk page warning had been given under this or his last account, and then suddenly the user was blocked as a major case, leading to dispute over the fairness of this, arbitration, and ban. I got involved because it didn't seem that the case history was clearly summarized and therefore it was getting heated rather than fairly considered.

(When I reviewed it, it became clear that there had indeed been serious defects in the prior handling of this editor, including a misunderstanding of evidence, complete lack of previous warning or other discussion on his talk page, evidence of willingness to engage in dispute resolution was ignored (and DR had never been requested), and his block log was clear under both accounts .)

The upshot was that SC was rapidly banned (3 weeks) for a year at Arbcom.

I disagree with this ruling on several scores:

1. The case should not have been accepted. The issue was lack of good case handling in the past, of a COI link pusher, not wheel warring or major difficulty (the contributors at ANI had actually sorted out that aspect already), and ANI can handle those very well. A comment to the efect: Reject - talk to the user, warn the user, and if the problem continues address the usual way, does not presently need arbcom involvement would have done the job just as well. The community is more than able to handle such actions by COI users. (Principle: Where the community can handle it with a gentle nudge, and no overriding principle is at issue to be decided upon, Arbcom should not as a rule take the case over, but allow the community to handle it and learn.)
2. Much of the impetus to escalate came from cross-wiki spamming. Evidence at ANI already suggested a question over this; much spam was misattributed and not the user's work.
3. Evidence tended to suggest the user generally showed civility and perhaps even willing to address disputes, but no DR had ever been requested. There had been no dispute resolution, nor warnings, nor an overriding "unusual divisiveness" amongst administrators at that point (WP:RFAR refers and a brief "reject" note would have explained this) - acceptance was contrary to arbcom norms.
4. The decisions focussed on SC's actions only; there was no mention of the community's failing to warn him, and in fact he had never been formally warned. (Principle: Even as a pusher and the like, arbitration should reflect fairly on all, even on long term "editors of concern", and not make one side seem more wrong than they are, or the other side less at fault than they were. All sides need to be advised if they could have done better, and cases neutrally summed up.)
5. Arbcom did not seem to investigate the background to the case effectively, or if they did the ruling doesn't show it - a piece of research that took me a couple of hours to collate.
6. The decision did not comment on any of the mishandling that had gone on that had led to this. For example, there was 1/ no affirmation that a user who breaches norms should be warned, if the matter is not so heinious as to deserve immediate block/ban. It didn't note that 2/ random contributors' comments at AFD considerable time ago are not sufficient to count as "warning", nor the corollary that 3/ users are expected to "get a clue" even if not "officially warned". These are all valid things that might have been noted, and reflect on the root of the case.
7. There was no comment regarding the apparent rush to paint SC blacker than was fair and to also marginalize well-founded concerns of evidence and fairness raised by reputable users, by some users at ANI, nor suggestion that there might be lessons to be learned for future ANI handling of "known problem editors". This is a very obvious finding, and a concern that is likely to happen again at some point, so a somber comment on good and bad case handling (and the need for the former) would have helped.

My main concerns therefore were 1/ the lack of fairness to the user, 2/ the taking over of the case from the community without real cause, instead of directing the community to fix the mishandling and treat it as usual, and 3/ that Arbcom must be willing to note if there is fault on behalf of the community, for many reasons (fairness, lessons, future reference when reassessing the banned user, etc).

Following the ruling, the old account was "blocked as a sockpuppet" (and added to "Misplaced Pages sockpuppets of Sadi Carnot"), which would tend to make SC seen as a user of prohibited puppetry in future. However upon checking, it is visible that Sadi had used the accounts completely in line with WP:SOCK (last contribs of WM to any article/talk/project page Dec 15 2005, first contribs of SC Dec 27 2005), and also, the old account was last used almost 2 years previously. I had previously pointed out these observations in compiling my "outside view".

There is no guarantee that SC would not have been rapidly warned, blocked, and reblocked longer, and so on, if left to the community. He may well have been, and the community has that power. The community mood these days is less tolerant of people who are unlikely ever to contribute positively to the project, if they cannot be brought into a more positive style. But that's fair, if he had continued. This serious decision, in the face of our mishandling as a community, was not.

Conduct RFC is a page that always catches my eye, even though I don't usually see a need to comment on it. The format:- Allegation, Evidence, Desired change, Certifiers, Response, and then the freedom for anyone in the community to comment on views or add a fresh view of their own, is one of our most open dispute handling formats. For that reason it often ends up one of the most likely to be fair; in and of itself it simply collates views, and tests consensus on different interpretations and views, in a very open manner that doesn't show obvious bias to any given person. The results show the range of conclusions reached, which is healthy - it's what one would expect such a forum to deliver if running well. Since by itself (and unlike ANI) it doesn't usually result in sanction, it allows the "silent majority" of fair and balanced users to show their views.

A possible weakness is in complex cases, or where it's hard to tell which of two sides is right or wrong (or both, or neither), but since conduct RFC is simply a forum for views, even then it's usually helpful, since extra views do indeed get raised, which help in future progression of the case. In a problem, sometimes just having others' input this way can help and reassure a good-faith user, that they have some backing within the wider community that they have acted well or their concerns are endorsed, and also feedback on anything they are doing that might best change.

To me, one thing thats highly noticable in conduct RFC is the way that even if earlier statements (or some statements) are non-neutral, as the RFC progresses, other more balanced voices tend to kick in and get support. In that manner, it's reasonably self-correcting, which is highly desirable.

Examples:

• RFC/Ryulong (July - Aug 2007) - one can see the wide range of responses, from many different angles. But notice how Ryulong has a good chance to justify his actions, or at least present evidence in that regard, and one can see two credible administrators concurring (demon, mackensen)... then a comment on the position of warning in block placement and whether the zeal was excessive... a concern that this was noted at RFA and was ignored then... a view on proxy blocking norms... a placement of the blocks in context and an assessment of where the problem might be... and so on. RFC is doing exactly what it should, and anyone reviewing the case has a clear idea the range of views and evidence, and how credibly they are endorsed.

• RFC/Alkivar (July - Aug 2007, taken to Arbcom Oct 2007) - A past miscommunication between Durova and Alkivar that led to concerns is part cleared up when the material is raised at RFC... an outside view that the case was raised maliciously or in bad faith... a "WP:COOL" haiku... a comment that everyone gets heated now and then, and that other admins are worse (which I note was followed up with request for details)... a comment that nobody comes out of the case looking good and the focus should be on improving the project, advising self-examination by the parties... a comment that takes up the diffs which disparage non-admins and criticizes it (correctly) as completely out of ordser and mistaken...

• RFC/Dbachmann (2) (Nov 2007) - A name I recognize from being asked to investigate the actions of User:Deeceevoice at Afrocentrism. One could not review that dispute without noticing Dbachmann being a regular on that article, as are users Deeceevoice and Futurebird. The RFC is certified and endorsed by 6 users, some of whom I recognize as being associated with concerns over POV pushing and edit warring. Despite its origins in a complex edit war with POV and fringe pushing involved, the comments seem remarkably suited to give an experienced reader (who doesn't know those involved) a good insight into the problems. The good and bad of all is presented and can be assessed.

It seems that by its nature conduct RFC usually does its job. Its job is not to "judge" or "rule" on cases, but 1/ to seek comments and gain insight from the community as a whole, what others feel, and 2/ to solicit independent inspection, and what uninvolved parties see as important in the case. In most cases it does this job very well. It is well visited, with a range of results suggestive of genuine openness, and insightful varied responses in most cases. As a result, an outsider reading a conduct RFC in most cases will rapidly get a good "sense of the community" and the different interpretations, issues and perspectives that exist. This is the aim of RFC, for the community, and for the involved parties on both sides.

The main source of harm (as with any process) is when RFC is used as a means of harassment or to misrepresent others or put them in a bad light. I'm going to suspect from what Ive seen, that in fact when this is tried, the community response usually is fairer than the harasser would wish. Nevertheless Im sure there are times it is abused, or goes wrong.

One exceptional case I got involved in was the banned sock-user DPeterson, who (as reported at this RFC, May 2007) had himself created an exceptional six conduct RFCs between October 2006 and May 2007, as part of his pattern of harassment against people opposing him in his edit wars . Most of DPeterson's accusations at Arbitration were untenable, spurious, and in bad faith; there is no reason to suspect these were any different. The cases were stuffed full of DIFFS and CITES. If RFC were easily misled then these would be examples to check, being 6 cases over 8 months. In fact apart from his own socks, not one of the six cases he tried to raise and accuse others of misconduct, gained support, and many gained no support at all despite the regular "sock-supporting-sock" activity from his own other accounts.

I'm sure that cases where conduct RFC goes bad do happen. But the format of the page mitigates against it strongly, and I'm not personally aware of any that did in fact go wrong as such. If you come across any, please do let me know - I'd want to take a look.

This answer is in four parts - general overview, personal aims and approach, arbcom handling, and summary, because the question touches on so much ground.

Overview - I have been handling the kind of edit warriors who end up at Arbcom since 2004. I took my first POV/attack warrior through RFC, mediation and (with approval of the ex-mediator and after seeking community agreement) to Arbitration and a topic ban, within 5 months of joining the project, and looking back the case statement was well drafted, and solidly evidenced. Since then I've dealt with several others. I discuss here the more problematic kind of edit warrior, the kind who may end up at ANI, RFC or Arbitration.

With any user said to be edit warring, there are many factors where judgement is needed, that influence the view, decision, and handling. Some of these are:

• Openness to change - Some are amenable to change, and just have a hard time getting the point, but are open to good support and advice if offered. Others seem not to be. So in many cases users identified by some person as "edit warriors" can be reasoned with, while others one ultimately concludes can't.
• Positive contributions - Some warriors have a record of valid contributions as well, or edit in other areas where they act well. Others seem to be editing just as another website to push their views on the world. (I've actually co-written an article with an edit warrior; many views and edits were extreme or unsupported, but his knowledge of the science of it was actually invaluable to the final article, even if it was difficult getting agreement on balanced wordings and removing some more extreme points.)
• Types of problematic behavior - Some have specific behaviors that if inhibited would address the visible problem without needing harsher measures (revert parole is one example).
• Virulence/nature of activity - There is a sense of virulence - the active harm done. Not all edit warriors are the same. Some just write tendentiously, but others are actively into personal attack, and more heavy duty warring. In some ways the ones who just grind down a topic are harder for other editors to deal with - user:Ludvikus on Philosophy (and possibly some behaviors described in the current Armenia-Azerbaijan 2 clarification case) come to mind - because they aren't actively attacking, they just make good editing impossible. (Which should not be underestimated either, it's also capable of being extremely damaging by driving off editors, stonewalling problem texts for months, preventing movement or collaboration, etc.)
• Honesty - Some edit warriors are 'honest' - they have their point or perspective, and argue it. Many articles have actually benefited from this range of perspectives, and the users have eventually been gotten to collaborate over time in some. But some are more dishonest - sock use, gaming the system, subtle mis-citing, refusal to get the point, and so on.
• Patience - Some editors may be seen as "not worth the trouble" by the community, and are referred to arbcom due to exhausted patience, even if their actions individually are not extreme ones.
• General situation - It's also a factor how complex the situation is. Some cases its one clear 'warrior', others its two 'sides' and their proponents, sometimes there is an antagonist to look at also... some cases are a near-hopeless train wreck of hostility with a few sane voices trying to stem the tide (a number of nationalist cases and almost perrenial edit wars come to mind).
• Effect on editorial environment - Last, although by no means a comprehensive list of 'factors', I'd put the factor that almost sums up the others - effect on the editorial environment. Our editorial environment is how we use the community we have, to write the content we're aiming for. So the acid test of any editor is to look simply at their impact on the editing environment.

I think that's a fair "first take" at factors to consider - their 'modus', honesty, impact, virulence, ability to change, and nature of contribution. Obviously each case varies. But as the above makes clear, there are many factors, and so it's to be expected each case varies. It's a judgement, and that's essentially what arbcom is empanelled to do - make judgements based on evidence how to handle difficult editors. Ultimately cases reach arbcom because the problem is one of the 15 - 35 a month requests (5 - 12 a month accepted in 2007) that the community feels can't be handled without a direct ruling. So with the understanding that no description can be exact, I'll try and describe my handling.

Degree of accomodation - We can accomodate (to a fair extent and if they agree) those who are learning our ways, and often use means such as parole and supervised editing, or topic restrictions to do so. NPOV and CIVILity isn't a stance that people always have initially, so some people take a long time to learn these and we aim to allow positive contribution without "throwing the baby out with the bathwater". But there are those who continually fail or show little sign of wishing to be a positive member of the community, or aren't able to employ the neutral point of view and respect of others' input which is essential for article writing, and in a significant number of arbcom cases the ruling therefore focuses on protection via removal rather than reform.

My general philosophy, (which I vary somewhat depending on the user's actual activity) is roughly - be civil and constructive, try to gain a genuine understanding and work with them, extend chances and choices - but at the end of the day the aim is still to gain fruitful discussion, and communicate the need for, and achieve, change, and to prevent further damage to the editorial environment. Editors who are not working collaboratively need to learn to do so (in whatever way we can), and work with others constructively and honestly, or it's doubtful there will be a place for them here. We're writing an encyclopedia, not providing general forum services.

Personal aims - In communication terms, I keep myself focused on what's needed to communicate and to listen, and what might help that happen, to try and work fairly with problem editors by starting with where they are at, and also balance that with clear explanation where things stand:

• for the community,
• for where I'm at, and the future decisions I may have to make (as a user with administrative responsibilities focused on the project's benefit), and
• for the user as an editor seeking to contribute.

I tend to always use both the "line clearly drawn" and the "helping hand" which I see as equally important points to communicate, but the emphasis and style depends upon the nature of their actions and harm being done, vs. attempts to contribute and not getting the point, and my own judgement what needs to be said. As an administrator my style is roughly, "help, advice and alternative approaches are openly there, but if these are not taken then this is where it stands and what must ultimately happen."

At arbcom the decision is going to be slightly different since one is not mentoring personally, but making a ruling as a reviewer at a single point in time, based on how they have acted so far and evidence provided of intent, conduct and activity. The emphasis is more on, what are they up to, what's the damage/benefit/prospects of their activity, what have they done with any help to date, and what is the view of experienced editors who are acting as arbitrators, on the user and the situation.

Personal approach - I've used both supportive discussion, and firm statement of policy, often together. As a result, I routinely can get the respect of edit warriors and "problem editors" and at times this has helped. Virulent edit warriors though often don't change easily if at all.

I look to communicate and keep trying to communicate all the way. If there is any reasonable hope for improvement, I'll be one of the first to try to seek it, and I'll discuss, explain, set out the "what must happen otherwise" and put my own time and effort into mentoring and supervising, make myself available... the works. I tend to be a good judge of people on this score. I do this myself and if it fails, I'll draw the line myself. Two examples:

• After the arbcom case on The Troubles, I got asked by one very respected admin if I'd mentor user:Vintagekits, a well known difficult editor, on the basis they thought he was likely to be banned but if there was anything salvagable, they couldn't think of anyone else willing to give a fair go who would lay down the rules to him and might get his respect. (If confirmation's needed I'm sure they'll confirm, as this took place privately Confirmed below.)

• Likewise I wrote the now-banned DPeterson this policy note, which combines a clear warning on policy and administrators' role, with the request even at that late stage to change. He didn't; and within 48 hours I blocked him for clear and further edit war conduct , then just over 24 hours later again for his repeat of the same edits when his block expired .
  
  (Other examples showing my approach in this area, are linked on my "additional notes" page here)

But my own personal tendency as an administrator specifically used to heavy duty edit warriors is, if I don't see that potential interest in good editing, or nothing happens as a result, is to tend towards some form of final option - a topic ban, or restrictions plus escalating blocks as enforcement by the community, or (if they are dedicated warriors of no notable project benefit), then removal, because we're here to get content written, not to babysit warriors, and if I think that's best for the project, that'll likely be the stance I take. An edit warrior with bad faith (socks, inability to enter genuine discussion, etc) I'm much more likely to conclude is a problem than a plus.

Knowing I have that approach in some cases, I make sure to listen to others' thoughts, if someone considers a lesser remedy for serious edit warriors is needed, but in those cases that I've stated at arbcom a site or topic ban was appropriate, my assessment that they would not reform was accurate:

1. Ciz tried to come back as a sock a year after his topic ban. When I identified the sock and called him on this, he next tried edit warring on unconnected articles, necessitating my opening a second arb case to seek an extension of ruling to cover unconnected articles and other shortcomings in the original remedy. I drafted the amended remedy for Arbcom (2005), aiming to rule out his multiple problem areas but allow editing in other areas. After discussion and explanation of the need, it was adopted word for word by Arbcom. No further problems have been noted since.
2. HeadleyDown burned out the arbcom-appointed mentors and was finally banned... then community-banned as KrishnaVindaloo... then blocked as maypole, jeffrire, and many other reincarnations, and admitted in email to doing it for "fun". Over a year later, I've blocked another four socks earlier this month (November 2007).
3. DPeterson, an unrepentant warrior to the last, evaded his arbcom year ban via an unblocked dormant account, to post personal attacks on his "opponents" talk pages, about their tax filings (of all things) and was community banned at ANI.

Arbcom remedies - Arbcom has several kinds of remedy:

1. Supervision of some kind - probations and supervised editing/mentorship are used where there is an intent to educate the user as they work, or positive work elsewhere. They require an assessment that this is an editor who does not understand, and if coached and watched may gain that understanding or "get the hang of it". It places them (or the article generally) under strict supervision in order that failure to do so can be addressed by administrators without delay.
2. Parole and editing restrictions - used as the term suggests, to inhibit problem actions whilst having minimal impact otherwise. These can be fairly flexible, being drafted them to suit individual cases, thus one might propose that a given process which has been abused (revert, AFD, RFC) may only be used to a limited extent, or with supervision... or that certain pages or sources of conflict are not to be edited upon... usually until decided otherwise, or until the user requests the condition be removed on the basis of subsequent good conduct.
3. Bans - which are the remedy most people think of first when arbcom is discussed. An editor is removed from the site, or from a topic or set of pages. Arbcom has set indefinite topic bans, but usually caps its site bans at a year. This isn't as loose as it seems; a returning edit warrior will rapidly find that their scope for warring is very limited (ANI and community ban took under a week for DPeterson, for example), or if they breach their arbcom ban, may find that the ban is reset or extended by the community to a community ban. Arbcom do not generally take it as their role to ban users permanently (possible extreme cases such as stalking aside).
4. Arbcom is not restricted in the types of remedy it may create, and can innovate remedies as it sees fit, in the best interest of the project. There may therefore at times be other types of remedy. The above are the most common.

Summing up - No one answer fits all warriors. There are a number of factors which direct how edit warriors are seen, including a large element of judgement and "reading between the lines", good faith where reasonable, and some hard nosed reality (we're here to write a project).

Arbcom must be willing to use any of the means in its armory. It must look carefully into the problem to ensure it really does understand the problem and has fairly assessed it, by using judgement to assess the nature of the war and those in it, their roles, and the remedies that might help. It must protect the project, which often means putting some form of protection in place such as probation, supervision, parole, topic bans, general conditions of editing... whatever it feels useful. A warrior who may make valuable contributions should be attempted to remove the problem editing without losing their positive input. An editor who can't change or refuses to, the burden may not be to any avail, and a ban may be appropriate.

The simple version when all said and done is that arbcom's choice of remedies is aimed at identifying the minimum measure that is likely to be needed to remove the problematic behaviors (with appropriate enforcement powers to address them), and making reform as likely as possible (if reform is deemed likely), whilst impacting as little as possible on any positive contributions that the user may make, or be capable of making, and giving the community a ruling that gains wide respect as well thought out.

My personal view on incivility (including personal attacks) is that I'm strongly against it. I've handled exceptionally virulent attack warriors and never found it necessary or indeed useful. Incivility and attacks are damaging to communal quality, and easily provoke some, dishearten or misrepresent others, and leads to escalation and bad atmosphere all round. It's also a serious breach of communal acceptable norms, not to be treated as trivial or minor. Further evidence of my stance on incivility and personal attack is on this post "Section break and outside view" on Swatjester's talk page (there had been a block of a known "difficult user", who responded with personal attack leading to an extension; I was asked to review it by another admin).

Editors aren't expected to be perfect or never have a bad week, or never be provoked. (Though personal attack is never acceptable.) Theres a range from curt, to bad faith, to snarky, through to vindictively nasty attacks, so there's a need for judgement how to handle each case. The aim is to avoid recurrence (rather than punishment) and protect the editing environment from the damage caused, which is the logic behind WP:NPA and WP:BLOCK endorsing appropriate use of blocks for these kinds of behaviors.

In general though, it is an unfortunate fact that incivility and attack do often repeat, so regardless of anything else and regardless of decisions in the present case, one should explain (or warn) the seriousness of it so they have a clear understanding where they may stand if it repeats.

Regarding incivility at Arbcom, I don't think Ive ever seen a user taken through DR and to arbcom for persistent "simple incivility" alone (though they have been blocked for it). Almost invariably it's turned into more problematic kinds of civility issue - edit warring, personal attack, bad faith and accusations, and so on. Arbcom regularly sanctions people who can't get the idea that you just don't do that here, with the twin aims of protecting the editing environment and hoping they will come to change their ways.

It's important though to check the circumstances; any attacks are "not okay" but some editors have otherwise good records and only end up at arbcom with attacks a factor because of (for example) some specific provocation, or because some bad-faith editor dug through their edits to find one or two that might be presented as incivility/attack. Others are almost professional provocateurs and tread the line to try and game the system. So one needs to check behind the dispute to get an idea what's really gone on. Sometimes one persons "attack" is another person's "assume good faith" (ie, a comment being taken the wrong way). But ultimately, if an editor is uncivil or the like, it's not okay.

How it's handled depends on circumstances. Some people are open to a ruling and probation; others will respond well to site or topic bans, (either short and escalating, or otherwise) ... some you simply block for an extended period, because their behavior indicates they are incorrigible, and (when they come back) if they repeat, let the community handle it thereafter. In general one tries to gauge the least thats needed to ensure the behavior doesn't happen, and if that means a site ban for a year, or a topic ban and mentorship for a month, or personal attack parole... case by case, best judgement. When the incivility was genuinely minor but clearly existed, it may be that it is noted as a finding ("X engaged in incivility ") but no significant sanction ("X is cautioned to avoid incivility in future"). That way it's not over reacted but nor is it ignored.

I'd always ensure there's a clear "what if" to any ruling though -- what if they improve (when will it be lifted) and what if they don't (when do we escalate or just remove them, or can any administrator deal with it). And I'd try to ensure the ruling contained enough scope that the community could handle it from there on.

Established editors and conduct standards generally - Problem actions by established editors are always a matter to consider carefully. As well as the usual factors, there will also be concerns about neutrality of the decision both ways (including perceptions about the neutrality and rightness of the decision), concerns about the appropriateness of criticism (by those who feel established editors should be allowed to get on with the job or given more leeway), and also the risk of losing a proven valued contributor of much experience if they respond by leaving. Much of Misplaced Pages's content work is done by a relatively small number of editors and IPs (a few thousand perhaps according to some stats), and so in real terms, whilst nobody is indispensible, we do want to try where possible to keep established and productive editors engaged and not disillusioned.

However, there is a hard line there. We have clear standards, and ultimately these apply to all. Over the years Misplaced Pages has now existed, not one policy has gained a section that notes established users have more leeway to breach conduct norms or a lower standard to meet - clear evidence that this view has not gained communal endorsement. We might seek to discuss or work around problems of this kind, but if the established editor (sysop or not) cannot or will not get the point, or their actions seriously damage their (or the project's) credibility, then action may be needed. We have one set of standards, and at the end of the day they apply to all, without exception. An editor who performs actions that are problematic, will have a detrimental effect - and the spread of perception that we have double standards would have a much greater detrimental effect rippling out from that.

By contrast firm application of our standards to all sends the message that this is our communal norm, which we respect and support, and a number of other editors will often recognize or respect that. (Note: we do not decide cases to make an example; what is being referred to is that a fair and well judged case decision that re-affirms or enforces communal norms with clear even-handedness will get noticed and a proportion of others will also adapt their behavior positively or feel confident asserting these as norms themselves, if they see this happen.)

Sysops - Sysops are users who are trusted by the community to have access to the tools, and not abuse them. They have respect as experienced users, but no other special status, and in principle all users are expected or able to make judgements as sysops do (only lacking the tools to execute some of them). These are old principles and communal norms.

There have been many discussions concerning the desysopping of admins. RFA is listed at Misplaced Pages:Perennial proposals ("RFA is our most debated process and nearly everybody seems to think there's something wrong with it, literally years of discussion have yielded no consensus whatsoever") and desysopping is almost of the same status.

I wrote in my June 2006 RFA that the primary difference between administrators and non-administrator registered users was, an admin has far fewer excuses. Specifically also:

"An admin has (ideally) been round longer, seen more, demonstrated knowledge and sufficient awareness of policies and attitudes to bear some responsibility for exemplifying the spirit and practice of the project. An admin also has more ability to do wrong if they mess up, to cause problems if they let personality and strong emotions get in their way, and to set a bad role model, and fewer good excuses of ignorance or acceptability if they do so. An admin has also represented to others that as a long standing Wikipidean he/she will be trustworthy to represent Misplaced Pages and help with certain kinds of decision for the project and has been trusted that they will help (and not hinder the project) if given that trust. They have also made a commitment to nurture the project. A non admin *should* do the same, but has not made any such explicit and willful commitment to do so, nor is Misplaced Pages at such risk if they don't, and therefore may have more rope. Another answer might be - although both are (ideally) trusted and competent, the minimum level of trustworthiness and competence is higher for the admin. And focussing on the role rather than the authorization, might be - a non-admin is on Misplaced Pages for the fun of editing and contributing. An admin is there for that reason, but also performs a function, which may involve mundane or a share of "chores", extra jobs as well as 'fun'."

Ultimately, all admins should enjoy a high level of communal confidence. Having earned it once does not protect against its loss later; it sets a presumption that the admin is doing the job well, but if there are visible serious problems or concerns, especially amongst their peers, then that is a serious concern.

General conduct of sysops - In terms of general conduct, administrators are not only judged on their trust with the tools. They must also have communal confidence, and the damage they can do by even ordinary misconduct (ie, not involving the tools) is potentially worse since administrators may be seen by users as representing Misplaced Pages and its communal stances and approaches, in ways that non-administrators are not. (See RFA statement above.) Thus an administrator who regularly engages in edit warring or personal attacks, and keeps on doing it (ie, despite warning or ANI discussion), may ultimately end up cautioned, suspended or losing their sysophood for this, despite the fact these actions do not actually reflect at all upon their tool use.

When the question of sysop conduct, and expectations of sysops has been considered by Arbcom, the committee has repeatedly ruled each time that administrators act as role models and will be expected to demonstrate a higher level of conduct. These rulings contain details of some of the expectations that apply to editors the community has approved to be sysops.

Reasons sysops end up being considered for desysopping - The common reasons would seem to cover issues such as:

1. Misuse of tools (deletion, protection, blocking in clearly improper circumstances)
2. Breach of basic policies (attacks, edit warring, biting/civility, privacy, etc)
3. Repeated/consistent poor judgement
4. Wheel warring
5. Failure to communicate - this can be either to users (eg lack of suitable warnings or explanations of actions), or to concerns of the community (especially when explanations or other serious comments are sought).
6. 'Bad faith' adminship (sock use, good hand/bad hand, gross breach of trust, etc)
7. Conduct elsewhere incompatible with adminship (off site attacking, etc)

Other factors that are not in themselves reasons, but contribute to the matter might include 1/ serious loss of confidence or trust, 2/ patterns of repetition or ongoing problems, and 3/ precedent states that administrators are expected to act to a high standard. (A factor in the other direction is that admins - like all editors - are not expected to be perfect, or never to make mistakes.)

Remedies and handling - Some of these are sufficiently serious that it is almost impossible to be an administrator if they are discovered. If this is not the case, then the case focuses on avoiding repetition. The principle is the same as for all arbcom cases - to keep if possible as much of the positive work, whilst removing the problems.

Admins are assumed to be able to listen, and take advice. So the default is that in all but serious cases, correction is sought. Three levels of handling of specific relevance to admins exist:

1. Cautioning - the matter is noted but no specific action taken at this time.
2. Suspension - a tangible sanction that is in a way, equivalent to blocking, in that it removes certain trusted access for a time. It sends a clear message that the conduct is firmly not acceptable, but also the message that although serious, the episode is over and closed once the remedy is complete (and that the admin is expected to change or loss of sysophood is likely).
3. Desysopping - used for conduct that is sufficiently of concern that it is incompatible with the retaining of the trust and credibility needed to be an administrator. Examples might include serious cases of: repetition or a pattern of ongoing problem conduct (especially if the subject of past formal notice at RFAR, RFC, ANI etc), bad faith actions (eg using tools for own benefit/agenda, forbidden sockpuppetry), and cases which would merit suspension but due to the specifics of the case or other factors present, is felt this is not adequate.

All administrators make positive contributions. The principle of balancing the keeping of positive contributions with removing the harmful ones, is broadly as stated above for edit warriors. We try ideally to retain their good input, and avoid their bad input. Warning (including caution or suspension) is usually an effective remedy, if it's going to work and is reasonable in the circumstances.

Wheel warring - Special attention is drawn to wheel warring. In many cases wheel warring is a once-off event for the admin in question. It is probably in most cases likely to be responsive to suspension, and/or caution that loss will occur if it should repeat. However in practice it quite often meets with direct desysopping, mainly due to its serious effect on the community, and its status as a highly forbidden act amongst administrators.

Note: - A factor not mentioned in the question covers conditions and processes (if any) to be followed for re-sysopping.

Arbcom needs to balance two things.

1. First, it should by and large respect the decision of the community, in forum, to decide on most matters. A community ban will often be based on a presentation and discussion at ANI, and are the prerogative of the community. To underline this, Arbcom itself limits its site ban rulings to one year only (exceptional circumstances aside). So the default is that where a ban is reasonable and plausible, the community's decision to act accordingly should be respected.
2. Second, Arbcom is (in effect) the place of final appeal. It should be open to addressing the cases where for whatever reason, natural justice, good faith, or the best interests of the project have not been met, or a defect in the discussion, relevant evidence or other factors mean that the ban has grounds for requesting to be reviewed. (A user cut off at arbcom has no other route to be heard except hoping some lone administrator will listen and take up their case. So rejection's a serious matter.)

There are a few cases where I think it's likely Arbcom should accept to review the ban, for example:

• The case was old, and the user may have changed, or states credibly that they wish to retry. We have a long standing communal view that nobody is incapable of change.
• There is evidence that suggests the ban, or events that led to the ban, or individuals 'pushing' for the ban, were not in fact balanced and fair. For example, sadly, we have had cases where an administrator has a grudge, and manouevered a problematic user to appear worse than they might have done.
• When there is little previous review, or the track record does not seem to fully support a ban. Some ANI cases, the user has a long track record to be examined, others there is almost no evidence except a few people's opinions.
• When the quality of consensus seems doubtful (very few people, not much relevant discussion, apathy at ANI, etc).
• When the evidence seems to suggest there was error or ambiguity, unduely negative assumptions, "pile-on" views or emotionality, or other exceptional circumstances, such that a review would be seen as fair.

In general a post that contains direct evidence that can immediately be confirmed to show "something questionable" was up with the ban, or a user with little prior record of concern, will stand more chance than (say) a user who has been at ANI several times and was banned a week ago for visible edit warring. Generally a ban should be easy to check if it's well founded, or if there is doubt, due to the discussion that should have gone on or other evidence that exists.

Arbcom should always be open to considering whether a case merits review, and importantly, if it rejects it, a very clear explanation of their rationale should be given. This is essential (and merits the time taken) since rejection by arbcom effectively closes the matter, and a user deserves the basis to be stated so that there is no doubt to the community or them, why this was felt right and what perceptions were relied upon in the decision.

Last, since community bans are creations of the community, I feel that there is a good case that where the user would get a fair hearing at ANI and ANI does not seem to be hostile to the idea (such as the first example above), Arbcom's role might be limited to considering the matter, deciding if it's fair to be heard as far as they are concerned, gaining agreement for unblocking for the sole purpose of discussion - and then direct them to ANI (which originally placed the ban) to have the matter reviewed by the community.

In general, where the community is capable of handling, and likely to handle, a matter fairly (or would do so with some initial guidance) then the community should do so. Arbcom is more for ensuring best handling of difficult exceptions where that might not be the case.

I've commented on the Allegations of apartheid case ((3) above). This was not a case that the case was dropped because the issue was no longer current; rather to judge by the RFAR proposed decision and closure, it was a case that had been mismanaged and was dropped finally because 1/ arbcom still couldn't decide what to find or say, and 2/ the matter seemed to be dying down after the nearly 3 months of no result at arbcom and could therefore be abandoned.

I try as a rule not to be rough and I may not know the whole story, but part of arbitration is to be honest with integrity. I feel this case was mismanaged and should not have gone as it did. So in a way I'm not sure what to say on this question, except.... case mismanagement and the failure to see the issue clearly, really only have one answer - learn from it, and don't do it again. I'm not sure how the first of those cases ended up as it did; the second of them may have more reason.

If it happens again... there are some basics that count:

1. Examine the conflict in depth, as a participant would. It takes time, I know this having done it many times. Sometimes it is the only way you get a good summary that everyone uninvolved can agree upon as a solid foundation for considering the issues and actions which took place, fairly.
2. Start with basics, and note the principles and main findings. This is possible in any case.
3. If the case truly is a problem, discuss it more and consider seeking wider views.
4. If there is still a problem, say so, and consider what is needed to at least help to address it. Don't make the community wait months, or have nothing back. Give some tools, some ruling, some help. No other group on the wiki can give a unilateral direction if there's a problem, and so arbcom needs to find a way on these to help the community if the community asks help.
5. If "do nothing" is right, then do nothing - but be clear why.
6. Be responsive and visibly active. It's easier to wait, and easier to bear a partial or disappointing ruling, if there's a degree of communication (which does not mean prematurely stating views!) rather than silence. Even just the act of posting proposals and support/oppose, signifies to a party that it's not forgotten; the arbitrators are thinking about it.

One further possibility that I haven't seen discussed or used, but might be useful, is also:

Ask for more and better evidence covering the point in question. The fact that we have an Arbitration Committee does not mean they will never need to ask the community if they feel more views on some point would be useful. This kind of situation doesn't happen often (and probably shouldn't really happen at all), but when it does, perhaps that's an option that might be discussed. Note that arbcom would value further evidence and outside views on the following, and see what happens. We're a community: the aim of arbitration is to accept evidence and make good decisions; it's not a court of law which is limited to what parties choose to present. Sometimes in the most exceptional cases, if the alternative is to fail to reach a useful conclusion, arbcom might consider (as an exception) stating that it would find further evidence or additional views on specific aspects, posted at /Evidence, helpful to a difficult case.

Case management - I don't think "all cases" is essential. There are many good reasons an arbitrator may not be involved in all cases, and good reasons (from a planning point of view) why we certainly should not assume that all will be. Obvious examples:

• Extended vacations (3 week holiday?)
• Medical or domestic issues (Broken leg? Marriage? House moving? Exam season?)
• Work (Year end? System upgrade? Short term secondment with poor connectivity? Travel?)
• Pacing oneself (an arbitrator may make the very sensible decision that it's better to take a month or two away from active involvement or step back to give advice but not be so involved at some point, than "burn out" or "fade")

The question also implies a look at how arbcom works. For example:

• Some arbitrators may do more of the work on one case for their peers. All arbitrators should be able to examine evidence and draft proposals, but on each case, not everyone needs to do this, or there would be significant overlap.
• In general no specific number of arbitrators are needed for a case; we empanel enough to hopefully ensure that all cases will be heard by a good number and when some fade a bit, the process is still viable.
• A fair bit of arbcom work comprises additional matters via email, and is not in fact "RFAR activity". So it is important to be aware that an arbitrator's visible activity may not match their actual activity. In general though, activity is measured by case involvement, as evidenced by posts, views, proposals and decisions stated on-wiki.

Opinion-forming and decision-making is a task which arbitrators cannot ask others to do for them - The bottom line, I think, is that what arbitrators are each needed for, is their view on the case. It might be enough if one arbitrator reviews some aspect of the evidence and summarizes it for the others to check, or one does the drafting (if done well). But each arbitrator who puts their name to a case proposal is certifying that on their own checking of the facts, they independently concur with that finding. Not that they have "rubber stamped" someone elses word on it, but that they themselves can on their own cognisance, endorse it as fair and accurate ( if a principle or finding) or appropriate and best (if a remedy). That is the crucial step that cannot be delegated by any arbitrator, to any other person. Anything else in principle, can be. They can decline to comment, recuse, and it's fine. They can ask others to sum up the evidence or check what user logs show, and its viable.

But they cannot ever certify a proposal as support/oppose/neutral unless themselves they have taken whatever steps are needed to ensure that they find themselves confident of it, independently, adding whatever caveats and observations they may have as a result of their work. (Otherwise we would have <N> parrots, not <N> arbitrators.)

Activity in all cases - Arbitrators have taken on a huge load, and should be prepared to carry out that responsibility if appointed. This doesn't necessarily mean doing all the work on each case, but it does probably mean shouldering a fair share of the burden, arbitration work on all or most cases save for clear periods of absence, aiming to provide long term service, and enough independent review of each case they are active on to determine independently their views, make proposals, discuss remedies, with insight and care.

They are also human, and will have periods of less activity, or none, for various reasons over time, and arbcom should be able to handle this. Usual courtesies should apply (discussing, letting others on the committee know up front, etc). If they are absent an extended period then extra measures should be considered, such as appointing an additional (or interim) arbitrator to cover their place, until they can reliably come back

(Detailed proposals covering "interim arbitrators", such as a "pool" of substitutes, have been discussed in the last while, but no formal proposal is yet confirmed; Jimmy Wales' ad-hoc appointments as he himself comments, have not worked well).

Summing up - As a result of the above I think the answer goes something like this: arbitration is a tough and demanding job, but it is what was signed up for and for the foreseeable future, we need people of suitable qualities willing to do it. It's preferable to have reliably engaged arbitrators who act responsibly and take defined periods off or a "back seat" which the committee knows about (or self-manage and sit in at the proposals and decisions stages or offer insight and such when real life is busy) than unreliable and patchy arbitrators, or burning out ones. Arbitration is a responsibility and we count on it communally; if the editor can't do it for a time, or faces disruption to their ability to give it fair attention, then I'd expect them to proactively handle this, manage their life, let people know as best possible whats up, and co-operate in helping the arbitration process and case handling be disrupted as minimally as possible. if it's a recurring or longer term problem I'd expect them to say so up front, and allow arbcom to begin taking steps to consider how to address it.

Note: I'm taking the question to mean "Should active arbitrators be expected to comment on workshop pages, and on arbcom discussion pages".

Short answer - Indeed. I'd like to see arbitrators generally doing more of that. Although the arbitrators' primary job is to actually address and resolve the case, posts placed by users on these pages express communal or individual concerns, views, proposals and feelings, and a response can be genuinely helpful and reassuring to parties. Arb cases are stressful for participants. A well-placed comment at the right time, and where genuinely useful, can help user discussion, which may reduce stress, and benefit parties and the editing environment. It is not time consuming or unreasonable to do so.

Purpose of workshop and discussion pages - The /Workshop pages are a bit of an unusual item; they function as a structured approach for participants to attempt to set out their views on the case and its handling in the same manner that arbcom itself would. They allow anybody in the community to set out proposals, and test the response, but they do not actually (in and of themselves) usually form the backbone of a proposed or final ruling. In general Arbcom decide those based on their own discussion and experience.

The name "workshop" is well given; they provide a venue for parties to trial their own ideas, which may or may not be used by Arbcom. These pages in general also provide a venue for individual parties to show their good or bad qualities more clearly, and as parties are warned, 'mooning the jury' and conduct during the arbitration itself are taken into account. Sometimes, in the well defined editing space of the RFAR subpages, they can provide the clearest quick indication available of which users are acting in what ways, and where the problem is, in a complex dispute.

Arbitration talk pages serve a similar purpose but for general case discussion, questions, non-evidence comments, etc. Whilst arbitrators time should not be taken up with endless lawyering or other re-enactments of the dispute which led to arbcom, when genuine reasonable questions are asked, they should be noticed by more than the same one or two arbitrators and more quickly responded to. Again, it's the visible evidence of arbcom being on the case, which will benefit those waiting for a ruling in these serious cases. A concern that users are talking into a vacuum for 4 - 12 weeks (as seems to have happened in some cases), or that key points are being overlooked in the draft proposals with no confirmation that it's actually okay or the point of concern has been taken into account (as has happened in others), is hard to see as a good working practice or helpful to an already stressed set of editors. We can always say to any matter, "Arbcom is aware of this concern and will comment after the case if still needed". It's obvious, but sometimes good for participants to hear.

Spirit behind these pages - The spirit behind these pages, and the principles involved, are important:

1. Communication by and with arbitrators is helpful
2. Feedback by arbitrators is also helpful. (Arbcom cases aren't a one way street.)
3. A structured forum for parties to have involvement beyond the strict discipline of the /Evidence page may be useful
4. The educative element for users of being able to approach a case as arbcom would, in /Workshop, has value
5. Comments, questions, and new perspectives and evidence may come up in discussion, which provide extra insight on the case, or on shortcomings in the proposed handling
6. Parties who see a variety of arbcom members responding actively on these pages will feel reassured that the case is being progressed, and will feel they may have some sense how certain aspects are being seen and where the case will end up, which reduces stress
7. Arbcom members who see parties responding on these pages will gain extra insight into the users, their patterns of conduct, and the dispute.

Mix of uses - It should be noted these pages are often a mix of genuine productive discussion, and forum ranting/debating. Arbcom cases often come with heated feelings; sometimes users are best simply allowed to talk these out while arbcom gets on with the real work of resolving and ruling on the case. There is a need to discriminate between useful and non-useful comments. But either way users should not be left in doubt that arbcom are aware. A simple comment "We are aware, thanks" might be all that's needed to address such a need.

Summary - Communication is important. The demonstration that arbcom members are active and forming views on the matters is valued by and very meaningful to participants. Arbcom members should seek to be visibly active on all or most cases they are active. It's usually easy, and if they are examining the case they will clearly have initial views on basic matters already. An appropriate comment on a talk page or at /Workshop might be stating the obvious, but even if they are "just comments", they can and do have a reassurance factor that's separate from any post at /Proposed or Final decision. Stress reduction is a Good Thing in and of itself, even if none of the /Workshop proposals were to be used in /Proposed decision.

Caveats -

1. The main caveat I can think of is that /Workshop is not the same as /Proposed decision: arbitrators comments are valuable, but they should be wary of appearing to vote on proposals there, or taking a premature stance on the matter. If a proposal has merit, it can always be moved to /proposed decision and formally voted there. There is a limit on what arbitrators can say without over-commenting on the case in hand, and too much risk of being seen as biased if parties proposals and viewpoints are perceived as being directly voted on by arbcom. But a note that they will comment on this or that point after the case is concluded, will achieve the positive impact and reassurance, and probably be seen as fair, without prejudicing the case itself.
2. If an arbitrator is visibly active on the /Proposed or Final decision, and no comment is needed on the other pages, then the aims of communication and visible activity are also probably going to be served, so far as case participants are concerned.

I don't know if it counts as 'proud', but the following list is a sample of recent article work from July - October 2007.

1. It is not a complete list.
2. The aim of the items listed was not FA standard but "a strong enough standard to be encyclopedic and solidly written".
3. Items in bold were authored or close to rewritten.
4. Details and diffs in detailed (collapsible) table below

Summary

I've answered (or will shortly answer) a fair bit of the "how should Arbcom look at administrators and should they treat them differently to non-administrators" (above (7) and - when posted - (12)).

I may well expand this but it's likely to be well covered by other questions already answered.

Take a look, and if you want me to expand on any part of that comment, do let me know, I'll be glad to :)

Why are nationalism disputes an issue - I could speculate several suggestions 'why':

1. The more people are invested in a stance on a matter, the more you'd proportionatelty expect to see them here (or on any forum). There are typically tens of millions of people in nationalism disputes. And they are already fighting without quarter for their view to win everywhere else, and on all other media. (Ditto religious disputes, territorial disputes, and so on)
2. A lot of the groups involved in such issues are relatively new. Many of these states and groups only gained a public voice in the last while (historically) and many are still gaining it, or harbor historic grudges. The need to massively demonstrate "I am here!" is non-trivial.
3. A lot of cultures and groups have little familiarity with concepts akin to NPOV. "Own point of view" is the norm, and in many places much more strongly than others. I suspect culturally that view is stronger in many ethnic dispute areas.
4. The "pile-on" factor, flavor of the month, and a bit of flash mob effect. Once word spreads, interested parties attend.
5. The usual logic "they'll find something else if ignored" doesn't work so well, when a number of the people involved have one core focus (their group's representation to the world), so the usual disippation of focus, drifting, "move on to something new" doesn't happen so easily. Some drift, but if you're into your Latvian/Estonian ancestry and ethnicity dispute, or Palestinian/Israeli dispute, or Greater Armenia, then no, you stay round to fight for it against all those who are doing the same basically, on the other side.

I do think that nationalist disputes are a phase. We'll handle them, because we'll learn how we want to handle them; then it'll be something else. Some will linger (see "perrenial edit wars") but things change, and I expect this one eventually will too. In the end, though, it doesn't matter "why". It happens, we need to address it.

Principles of handling - Treatment proposals vary as widely as people. The basic principles described in other questions (eg, #(5) edit warriors, #(6) uncivil editors, #(7) use of desysopping and #(12) should some editors be considered 'more equal' than others?) hold, and it's important to recount those before addressing such a heated area, lest we overlook them by accident:

1. All editors capable of working productively with others (and not doing harm) are welcome to edit.
2. Many of our stable neutral articles got there due to strong views hammered out the hard way, and lessons learned. As well as articles, entire processes, adaptations, and approaches emerge from difficult problems. (Misplaced Pages is a learning community, we expect to have problems, that's how we learn, if we're good at it.)
3. Core communal conduct norms (No personal attacks, no edit warring, etc) are ultimately, like NPOV, non-negotiable. We expect civility, good-faith actions, non sock use, no attacks, working with others collaboratively, respecting others even if views differ, and so on, and editors who consistently can't get these ideas (despite all our efforts to help) may not ultimately have a place here any more than those who can't get the hang of NPOV.
4. Before diving into extremes, we try and accomodate, and we try to keep and encourage the good, whilst deterring the bad. We already have a great toolbox of ways to help this happen.

In addition it's important to remember, we can freely innovate.

Handling these disputes - What seems to me is the case is, we have the tools we need; we may not yet have worked out a consensus on using them and editing within these kind of disputes. They are part of a more general group, perennial edit wars, which we need to consider how we'll address in future. It's unrealistic at present to try and have a credible article that reacts like a snake held by head and tail, and hope it to lie still. We have a wide range of perennial edit wars, a number of which predate the nationalism ones, and rather than focus on the nationalism ones only, it is better to look at a solution to the kind of article that inherently will attract dedicated warriors, and ask how that kind of article can be brought into line, given stability, good editorship, neutrality, collaboration, even though we know it will attract this kind of attention for the foreseeable future. There are many obvious ways, each of which has pros and cons.

Commonsense says if an article attracts this kind of user on a regular basis, then we need a means which helps more rapid handling of problem editors without many months in dispute resolution. If dispute resolution is going somewhere - good. But often it is used as a stalling technique. (For example this comment by a virulent edit warrior and sock user.) That's an example of gaming: for example, mediation should be where people discuss how to accomodate all significant views, not a means to stall progress. What we ideally want is to make the debate productive and able to reach a conclusion, rather than kill it dead - to not lose the benefit of these debates, in trying to resolve them. (Community enforceable mediation is one such attempt.)

Established norms may be up to the task, used appropriately and well - My gut feel is, established communal norms that exist, are actually more than adequate. We may need to learn how to apply them, or have to "grasp the nettle", but we have highly supported norms which can, in principle, and used with insight, handle such issues.

To pick one example of how existing norms might be useful: I have noted many times when working on pov warriors, that edit wars actually require considerable attention and involvement by the warrior to dominate a page. It's really hard to run an edit war if the substandard conduct might result in even short-term loss of access to the page concerned. Two 24 hour blocks on DPeterson for re-adding points against concerns of others and failing to respond to others' concerns, then re-adding again after the block expired, were perceived as a welcome sea-change on the Attachment therapy edit war, during arbitration. That was all it took to 'break' the actual edit war there which had been running a year by then. This suggests that one approach to perennial edit wars may well be require a more complete level of policy/norm compliance on certain heavily warred pages, failing which remove editing access briefly. (eg, keep blocks shorter but be willing to use them more within edit wars for repeated low-grade tendentiousness, too.)

That may be all that's needed. Perhaps on those pages, if we brought a number of more experienced users into the frame (rather than just leaving it to existing parties), and we also insisted upon real NPOV editing, good cites, no attacks, no incivility, no stonewalling or other games, no lines of discussion that are evidently intended to subtly prevent progress, users willing to learn when something's not okay... and we accepted sincere earnest editing, aimed at encyclopedic content only... perhaps this would make edit warring harder. Maybe one remedy or ANI decision we might reach more often is to endorse a stricter editing regime regarding editors' constructive involvement on these articles. Maybe we could also make it easier to obtain closer supervision of editing on an article, by deciding it at ANI instead of Arbcom. Bans are routinely handled at ANI, perhaps ANI might also be used for the community to decide to establish a supervised editing/probation regime on a problem article, without needing to route it via Arbcom? Perhaps we might restrict all users to the talk page, place the article under the control of a group of experienced uninvolved editors for a month, and they rewrite it based upon talk page discussion, to set a direction free from interference, before handing it back. Draconian, but would a year long edit war driving away others, be any better? Maybe if we remove or handle the warriors, their place will be filled by non-warriors. At the least, good quality contributors will be less deterred, and less driven away, or may come back.

There's many approaches the community might experiment with; some would work, some really wouldn't. This would have to be a community decision, not a sole editor's or even Arbcom's, and would probably emerge as a practice over time (gaining 'consensus by vote' on such would be difficult at best). But right now, just noting here that we haven't even begun exploring options how such wars might better be handled..... that I'm happy to do.

At present it's remarkably easy to edit war, as witness the current Armenia-Azerbaijan 2 clarification case. According to one comment, all we have allegedly achieved is that our nationalist edit warriors are merely trained to be more civil when they war(!). Also various other "slow motion" or stalled edit wars. It is even now, remarkably difficult for most editors or administrators to take action against someone who is merely unhelpful and unco-operative, rather than outright hostile and flagrantly in breach of policy, or to justify removal of a borderline editor who does no explicit wrong, and especially if (like many difficult editors) they make occasional good edits. Perhaps we need some way to address edit warriors who merely make matters difficult or tiring to engage. Again, this disruptive behavior would not require a new norm to handle, merely a new appreciation of these norms. Flagged revisions (when implemented) may mean that warriors can war all they like but can't get their versions in front of users. Lastly, we may simply get better in handling them just by passage of time.

The bottom line - The conduct principles that we must always come back to though, are those listed above and in answers elsewhere. Articles belong to the community and are written by them, not by any one group... anyone (with limited exception) is welcome to edit... collaboration and collaborative "buy-in" on decisions is crucial... heavy-handedness and sanctions are not really desirable except in clear and obvious cases where they help protect the project, in exceptional cases, and when other routes have been tried and failed... seek dispute resolution... keep NPOV... and so on. For every edit war there are literally dozens or hundreds of articles with broad consensus. The bottom line is, we have community principles with immensely high buy-in, related to communal management and authorship of articles, communal consensus (and many eyeballs), and communal acceptance of diverse (often strong) viewpoints which may clash or come together to find a neutral point of view. Throw that stance out or get too one-sided in the rush, and we might as well all go home. The rest is an inconvenience and a problem. Even now, we're surviving these wars. They're intensely irksome, but not fatal, and we'll learn. By contrast, a general agreement by the community upon NPOV, valid research, and that even if tough, we endorse communal open collaboration in presenting these, is Misplaced Pages.

I don't claim that any of these would be 'right' or would magically fix it. I suspect edit warring is a long term problem. These are simply examples, showing we do have approaches we haven't tried yet.

I think everyone can think of some item they would like changed or improved - at least now and then.

If I feel a policy genuinely would benefit from changing or introducing, I tend to do it, or ask for communal consensus to do it, rather than ignoring it. Policy changes - like much of my work - are prompted by some thing that makes me realize a policy is not completely doing its job and the community has not drawn correct understanding from it. A list of my recent policy page work is given below.

If a proposal isn't accepted, it usually means the time isn't right. Policies are the community's work, and added to help the community, so that gets respected. However, I've generally been successful at choosing changes that the community endorses as sensible, and wording them well, so in fact a sizeable majority have been accepted.

So as a editor familiar with policy development over several years, my first thoughts are two possible answers, and they are tempered by experience and forethought:

• There are many things we could improve. But policies and practices evolve by finding what works and what doesn't, so any change must carefully weigh up the potential to be on the whole, unproductive or counterproductive (the community sanction noticeboard's demise was one such case, for example).

• Likewise, we could write better content by changing our ethos on many things. Citizendium for example tackles socks and trolling by removing anonymous access. But our ethos is something important to us. We value it (as a community), we have immensely strong buy-in to many parts of it (and the principles backing all core policies). We have these even when they cause us problems (eg not instantly banning all problem editors), because we communally have made the decision it's what we feel will serve best in the long run. So ultimately, we keep many practices that we know cause us trouble in some areas, despite the degree of difficulty that the decision will engender, because communally we believe they are right ones for the project, even if they do sometimes come with a price in terms of stress or headaches. So any change must consider not just "what would functionally improve the project" but also, what would fit culturally with the ethos (or spirit) of the project, too. We are a community that produces an encyclopedia, and our culture and community, varied though it is, is an integral part of that.

What these mean are, changes to policy are best to arise organically, out of editors actual experience, than conceptually.

I've been involved in an unusual degree of core policy editing; if I was aware of a policy that genuinely needed improvement or change to better reflect how we work, I'd be considering whether to raise it for consensus. I tend to work on policy when I notice in passing that editors are facing a genuine problem and that a problem with a policy page was part of that. As such, my policy work has included fixing of points prone to misunderstandings or misuse (RTV, BLOCK); clear wording, explanation and focus (NPOV); howto's (BANAV, GOOGLE); explicit attention to loopholes used by edit warriors (GAME, NPOV, POINT); core pages that are widely viewed (ABOUT); pages where a community concern is evident (BLOCKTEXT information); and other useful information which is of confirmed relevance to editors.

The following are some of the main ones in the last few months (full list here). Note that as you'd expect, most of these are small improvements rather than drastic ones:

Page	Modification	Motivated by	Date	Before / after
WP:CEV (Confidential evidence)	Proposed following "sense of community" - Use of so-called "confidential" evidence as the basis for blocks and other allegations.	User discussion took place at WT:BLOCK in the wake of the Durova/!! affair, which seemed to be calling for a proper policy rather than a subsection. It's still under development, being new from scratch in the last week; time will tell if it is used or rejected.	Nov. 2007	new
WP:COI (conflict of interest)	Nutshell rewritten to be more useful and less BITEy, and actually sum up what we want people to do.	OTRS 'quality' queue email, which I would have directed to WP:COI except the nutshell was too 'BITEy' and didn't in fact explain COI when I went to check it was suitable.	Oct. 2007	talk diff
WT:NPA (no personal attacks)	Following tentative agreement by the community on a resolution to the notorious BADSITES case, proposed the final edit adopted to fix WP:NPA.	A BADSITES draft consensus was reached on WT:NPA - but as an outsider visiting the decision, it seemed the policy already contained the seed of the proposed addition and only mild clarification rather than extra section was needed.	Oct. 2007	diff1 diff2
WP:RTV (right to vanish, 'meta')	Authored. Approved at meta as the current stable version	OTRS issue, regarding common problems, misapprehensions and flawed expectations of RTV, plus deletion of evidence, and the need to restore pages if users return. RTV policy was a poor explanation.	Oct. 2007	before talk after
WP:WBE (Wikipedians by edit count)	Rewrote caveats to partly address the WP:MFD concerns on this page	Page had been listed at MFD due to concerns over "glorifying edit count", seemed that part of the problem was it didnt explain itself (or edit count issues) firmly and well.	Oct. 2007	diff
WP:BLOCK (blocking policy)	Correct definition of "indefinite block" drafted and (after some discussion) added	Noticing emotionality/stress in some block case, which had arisen due to policy not explaining what indef blocks are actually used for, leading to the very incorrect (and widespread) misunderstanding that they are equivalent to banning.	Sept. 2007	diff
WP:N (notability)	Added important definition of "presumed" which made WP:N policy consistent with WP:NOT (sourcing is evidence but the existence of sources alone is not conclusive - announcements, transient media, and so on being examples. WP:N gave the impression that sources were enough, leading to misunderstandings and incorrect contention at AFD.) Also cleanup of footnotes.	AFD case in which this important qualifier was not understood by one party, leading to endorsement at DRV and the observation that the guideline used the term without explaining it.	Sept. 2007	talk diff
Arbitration policy/Case handling	Authored this arbcom information subpage.	Hand-holding and supporting arbcom parties through multiple cases where there were long delays, attacks etc and no clear explanation why, leading to avoidable stress for users.	Sept. 2007	new
Mediawiki:Blocktext (User interface, block message)	Resolving the mailing list concern whereby blocked users were not given good directions regarding email and appeals. (I'd previously written much of WP:APPEAL)	wikien-l (email list) discussion of unblock problems, whereby users were told to email admins who might never reply or might not have email set, before using the unblock template. Seemed to me that the problem was the page should say both are options, not "one then the other".	Sept. 2007	Discussion
WP:BAN (banning policy)	User not to be described as 'banned' without good basis for it. After various discussions, policy changed to add "and the block has received due consideration by the community" in two places.	Some ban review or other, where a user was described as "banned" when in fact the ban was only being discussed or pushed for. The premature use of the assertion in such debates to imply the community has already decided, might have a pejorative unfair effect due to "crowdthink".	Aug. 2007	before proposal current

Prior to this I also wrote two pages in July/August 2007:

• WP:GOOGLE, a comprehensive guide to search engines and search engine tests on Misplaced Pages, motivated by problems in Google use seen at AFD,
• WP:GAME, prompted by many serious disputes where good-faith parties would have been much better able to argue edit war cases at dispute resolution or ANI, given a good description of gaming which would let them say with certainty, "user X games the system".

and also previously, wrote WP:ABOUT and de-crufted/restructured WP:NPOV.

There are some things I think would help us to look at. But we have a complex community with a lot of interacting and fluid aspects. 'Magic wand' answers might end up just a bit too much like the failed community sanctions noticeboard or Esperanza for comfort... they might help us one way but prove a year later to harm us another. Innovation is important, but I wouldn't actually put forward a novel idea without a lot more thought if it genuinely fits with what we have, meets a need, would be valued by the community once implemented (ideas that are commonplace now such as checkuser access were somewhat controversial when broached), and above all, where we as a community want to be.

There are many ideas that would be nice. Experience of what has worked positively on policy editing, though, suggests strongly that 'measure twice, cut once' is how I prefer to do it.

FT2  17:36, 1 December 2007 (UTC)

A good question.  If it's okay, I'll answer the question itself, and also answer directly, the implied question behind it, which is about handling of one's own views in a conflict, and open-mindedness to others.

As a rule, I like to use a combination of commonsense and communication, which surprisingly often, others find agreeable. I tend not to jump to conclusions on matters I don't know about, and keep a level head. So a lot of communication is about being able to realize and state what's important to say, draw lines if needed, but also having a sense of reason and perspective, and listening to how differences arise.

I spend a lot of time working on disputes. I handle conflict and disagreement well enough that my RFA nomination was a result of my approach being noticed in exactly the kind of matter you describe, one where I had a strong opinion, and was met with strong opinions too. It's hard to be sure whether this is useful, or "blowing ones own trumpet"; however here is the nominators comment, and a couple of others, relating to how I conduct myself in a heated discussion:

"cares about the project, does not mince words, is proactive in finding common ground with others, does not hesitate to give credit to his fellow editors when due".
(Jossi - Jan 2007 - RFA nomination)

" overwhelming degree of sanity and levelheadedness ... has never lost his cool".
(Premeditated Chaos (PMC) - Jan 2007)

" few admins who could work with despite his flames, and also understand the need to protect the community".
(Alison - Nov 2007).

So in a way, it's hard to pick out specific times I "changed my mind" because most of the time I'm listening for what's going on, and how that affects the situation. If one thinks clearly and is honest, then it's all about exploring what's best.

That said, here's one AFD nomination where I have changed my view to recognize a possible consensus during the course of a debate (sort of), and one article disagreement which is almost a classic answer to the question.

1. AFD - Kevin Eggan

Appeared to be a "puff" bio mostly, on a bioscience researcher who has some minor press notice, one award... and everything else both online and in the article was essentially expressions of hope of what he might one day do. Very questionable lack of substance. Notably, its author was also caught running "pay me and I'll get you a Misplaced Pages page" type adverts on Ebay.

On reviewing the article, historic notability (WP:NOT) seemed fairly uncertain, WP:PROF wasn't evidenced, and when examined in depth, most of the notability claims seemed to have a disturbing reliance on WP:CRYSTAL. The article cited papers, but nowhere seemed to document if they were important (having written papers doesn't imply notability either). It could hardly seem more one sided. Having solid reasons for skepticism and little/no real evidence of anything notable that stood up to checking, I nominated it for possible deletion.

The AFD reception was mixed, with emphasis placed on the nature of the award given. I felt that emphasis was being given to what was actually for the most part, just well presented smoke. I wasn't convinced, but by the middle of the AFD it began to look like there might be a consensus forming anyway, or a perspective on his award, and so on. Or at least, it might get kept.

As a result I decided the right thing would be to revisit the article and rewrite it from scratch, further research and add missing information, and focus on his present work rather than speculative hopes, and see if that would help. The reason being, he had not seemed to have evidence of notability on checking previously, hence not worth a rewrite before AFD, but if the community would decide he was notable then a rewrite became essential to at least refocus the article on his actual work as a proper biography for future, not just puff, media soundbites, a papers list and a speculative future. Even if borderline kept only. So I did. before after. The AFD was finally closed as "no consensus", which signifies "keep", which in retrospect is an appropriate view. The current article is visibly focussed on past and present, rather than speculation, and is more obviously reasonable and valid content as a result.

2. Article - Casino Royale (2006 film)

One clear case where my mind was changed in the course of an article disagreement was the plot summary for this film, between March 26 - April 8, 2007. Because it's spread over four pages, I've summarized it here:

- I edit the plot section, which is missing some serious gaps and therefore doesn't actually explain the plot well.
	- Bignole full-reverts the edit with the curt comment "Wiki is not a substitute for watching the film". I disagree with both the opinion and the full revert - a quick check of the previous edit shows that the changes were small, select, reasonable, and focussed.
- I drop a note on Bignole's talk page explaining that the edit was small but crucial details, and that his revert has 1/ left the article not explaining the film, and 2/ in general blanket reverts are a poor idea since they give the impression the edit was not in fact reviewed. I ask him to re-review the diff, and reconsider.
	- Bignole responds, citing 300 extra words, and excessive detail added, and ends by suggesting "Please read Misplaced Pages:What Misplaced Pages is not#Misplaced Pages is not an indiscriminate collection of information and Misplaced Pages:WikiProject Films/Style guidelines#Plot for a better understanding of what should be in the plot."' A minute later Bignole reposts, explaining: "Also, the blanket revert was not an attempt to make it appear that your contributions are not appreciated, only that this plot was trimmed from an excessive length by several editors months back."
- I post and ask Bignole a second time to review the original diff. I state that "There are some significant plot jumps in the original. I take your point on what is and isn't excessive detail, but don't agree a 100% blanket revert was the best way to proceed", and cite examples of these.
	- Bignole responds, this time beginning "Plot should not explain everything to someone" and following with a large number of carefully described examples of style and detail issues. I decided that he's almost certainly right, and learn a fair bit about ruthless conservation of words and better focus on film plots, which stands me in good stead for future work. I apply the lesson in future, earning comments such as this edit summary on later plot edits.
- I re-edit, taking note of these points, to fix the plot gaps in a style and approach indicated by Bignole's comments. - I also drop a brief bullet point list on the talk page, listing the 11 edits, so that if there is a dispute we can discuss them point by point. - I post to Bignole's talk page, "Generally, I agree with your comments on plot detail and length of section (thanks). So I've kept to a minimum this time, and to make it easy have summarized the edits on the talk page in a short list." In case he isn't happy though, I also state, "Please let me know your comments, but can we discuss rather than revert if you have concerns, since this edit looks good to me."
	- Bignole responds, "I went through it. It's fine. I didn't revert anything so much as I just reworded a couple sentences." - Bignole responds to talk page list, noting "I'm fine with most of it" and that a couple of things have been reworded for further improvement, which I'm happy with.

Answer amended Dec 3, to clarify possible ambiguous points.

Unusually, this will get into personal reasons rather than on-wiki views. I'm okay with that if you are. I stated that see myself as accountable and I value transparency, and this is part of that.

If the main question is whether this is some kind of afterthought, then no. There are several people (including 2 crats, a few checkusers, and a number of other candidates) who I'd discussed it with and knew it was confirmed well before nominations opened, and who know the reasons for it. Several others had wanted to ask as well. If you need confirmation let me know, I'm sure they will.

In terms of effect, unless a user literally self-nominates with a day or so to go, then it's probably either neutral, or slightly adverse, since it would remove time for answers, and cause absence during the early days when attention is more widespread. This is easier seen if one looks at the aim of this process - to find the people that are felt best to fill the roles. A period for questions is part of that. As it happens, in this case 1/ most people's "stock questions" are drafted and can be repasted, 2/ individual questions and examination of track record can be drafted in a day or so, and 3/ even 7 days is considered ample for other searching examinations and discussion of candidates (RfA/RfB). A candidate who nominated (hypothetically) on November 28 would still have 5 days of Q&A time before the election and probably be fielding questions for several days during the 2 week election as well. Perhaps one might shorten the nomination period to 26 days, and then put a week (rather than 3 days) between last nominations and first votes to guarantee at least that much time for questions and reflection, but overall, I don't think it makes much difference in practice. Users who want to assess a candidate are still very much able to do so, users who don't, won't. But for the actual reasons, see below.

Two main factors: -

1. I had emailed the OTRS list early October (Oct 3) to comment that I had visitors for 2 weeks that month until Oct 20 and would be unlikely to be able to do much OTRS work in that time. For example whilst Aug 22 - Oct 4 averaged 20 - 50 edits a day, Oct 5 - 16 were around 5 - 10. After catching up on matters arising in that time, I arrived at early November with an exceptionally unusual 3 weeks behind, and made the decision to prioritize four serious matters. One of those matters was an email correspondence with Jimmy Wales related to the general future of the project, that I'd promised him a reply on, and another was three well known disputes that had all been through WP:ANI and needed careful addressing: the AWeidman and sockpuppet community ban issue following the user's renewed use of personal attacks (AWeidman, DPeterson), another was verifying and blocking the latest sockpuppet ring of community banned vandal HeadleyDown (Spoctacle, Realbie, Lingorama, Arlen Wilps), and another was the Chidiac dispute (Achidiac, Rdpaperclip, T3Smile). These can be seen in my logs for Oct 31 - Nov 5, although some have continued to need dispute resolution off-wiki via email with the users. I was also wrapping up several peoples' help and advice. It seemed right to conclude these previous commitments before engaging in new ones. Alison and James F will both remember the volume of subsequent dispute resolution work by email, between myself, Chidiac and Rdpaperclip, following their blocking, from having sought a 2nd opinion on the matter, for example, and the care needed due to the volatility of the case and the related legal matters.
    
2. Secondly, I like to write neutrally and fairly, which needs exceptionally care when writing about oneself for others who may make a decision in part on that basis. Case insights and judgements are easy. Writing about oneself takes - if a person has integrity - a certain additional degree of reflection, consultation, and caution.

An ability to choose priorities is important, and at that point, I felt that dispute resolution I was involved in handling and advising on, had to take priority over elections, since the former are part of my present role and capable of present disruption, and therefore a present responsibility to the community.

Concerning myself, I often apply a stricter standard (as witness my first RFA's withdrawal comment). Apply to Arbcom... it simply took time to bring up to date the pages others might reasonably expect to be able to check, relating to article and project work, and a statement that I felt comfortable with the wording and that it represented me reasonably, with evidence diffs for statements made, which in turn required locating. I caught up on the outstanding dispute management around Nov 5, the user subpages around Nov 11-12, the email to Jimmy Wales on Nov 14, was away the weekend 16-18 (booked months ago), and then was free to attend to the arbcom statement, posted 3 days later.

If you need further detail on anything, please do let me know.

Last year I didn't focus on the arb elections at all. Like RFA, I figured it had enough eyeballs, and I pretty much as normal prioritized the focus on the range of work I was doing (dispute handling, content, and so on).

To speculate now on whether it could have been predicted assumes an involvement in that election which I didn't have. I could make a comment, but "it was/wasn't clear in hindsight" is usually subject to the rule of thumb that hindsight has 20/20 vision.....

A better question is to ask 1/ what might help in future, and 2/ what the arbitrators themselves think of it.

Overall though, since I wasn't involved in their choosing, whatever I might say right now would be of doubtful resemblance to the reality of the statements, comments and questions which faced the community at the actual time. I can make a few guesses though.

Arbitration is a tough role, and makes demands in many areas. Examples:

• Arbitrators will need to be highly experienced, have good insight into the community's spirit/feelings/norms, be tenacious, willing to work hard
• Arbitration work is demanding in time and energy. Cases are intense and often thankless, there is an unknown amount of private work too.
• Users have a private or working life too.
  (Which may also be routinely or unforeseeably busy at times)
• Arbcom burns out even experienced users.

My guess is that some of these are easy to see. Credibility, experience and "the right words" are easy to see. But how deep is the inner determination to ride the wagon if appointed, and take it in stride? How well judged is the unknown workload against personal abilities?

The ones who are still active a year or more later... we need to identify those a year before. And there isn't an easy way to do that. Maybe we could make it easier, but really we're looking for exceptions, people who are not only excellent as editors and dispute handlers, and respected, but especially, people who can take this kind of workload and handle it, and can keep doing so.

People have real lives... it's easy to forget how demanding cases can be, how much extra work goes into the role, that even arbitrators may not always find working together easy, and how long Committee members handle it for. It's possible some do a lot of background work, but arb cases is how the community gauges their activity, and a major role, so it's a fair means of assessment.

I suspect a lot of the patchiness is because an election process foregrounds somethings at the cost of others, and even those who are close to someone can't always read this kind of question well. We have people of credibility, experience, impeccable records (when appointed), widespread confidence (85%+ in 2006). But we don't have the same sight of how they'll handle life once appointed. And "popular and can do the job" does not always mean "will handle it from a time, effort, and focus perspective".

I think that's what's missing - and the thing is, it's not obvious how to gauge it.

(And if we do ever work out a way, it'd make a fortune in executive recruitment...)

This was a common question last year, so I disclosed it in advance this year. It's already out there, at the top of my nomination statement, under 'real life info'. And to the second part, yes (at present and for some years).

I manage my focus, and part of my management is selecting what to focus on, on the wiki, to ensure that I'm active where I can be of most help. (For example, I'm active on serious disputes and 2nd opinion work, because these confound many administrators so a clear analysis can be exceptionally valuable, but I don't patrol RFA which is well attended by many equally capable editors.)

Having undertaken heavy duty dispute handling and examination at others request, and ruled on a great many complex disputes and taken several virulent and difficult warriors to arbcom, I have a very clear understanding just what to expect such cases to be, and a fairly realistic experiential idea what I'm getting myself into. I'm virtually doing arbcom-style casework already, in all but name, and have been doing these tougher cases a long time (since 2004). I've also been tracking this highly sneaky and virulent vandal's reincarnations since 2005, for the project. If you look into the evidence in Attachment Therapy (2007) for example, which was a very complex case of sock evidencing across 7 accounts where checkuser failed, it's thorough, self explanatory, and "tight". And obviously, a lot of work. So its a workload I'm very familiar with. It'll be tough, but managable. The latter is what counts.

By coincidence, I have an admission to make here. I was the user who worded the description of SPOV vs. NPOV, and (after gaining consensus) added it to the NPOV policy. That was 1st August 2007. So I would hope I can give a good description of these for you now.

Involvement in NPOV policy development - I've been a significant policy editor on WP:NPOV for 18 months since June 2006, and in that time undertook its refactoring into its present structure, together with several wording edits, and section removals and additions (it was in quite poor condition back then). I also spent much of late 2005 in talk page dialog and dispute resolution with the now-banned vandal warrior HeadleyDown who POV warred for destructive goals on the basis of (spurious but intense) scientific skepticism. So my understanding of our policy on NPOV is probably a good one.

The focus of this edit was to improve the wording of "list of biases", to note that when a pro- or anti-scientific viewpoint becomes so excessive as to no longer treat notable viewpoints neutrally, then there is a problem. The existing wording had also omitted anti-scientific bias and scientific skepticism, which are essentially under the same umbrella.

See original edit.

The only criticism when I made this proposal was that I'd used the term "scientific orthodoxy" (technically a correct use of the term, to mean "that which is accepted as mainstream by scientists") whereas consensus felt this had negative, religious and dogmatic implications and felt a change to "prevalant scientific opinion" would be a better term, which I agreed.

This version was then agreed, edited into policy, and remained in WP:NPOV untouched from Aug 1 (when added) until Nov 4 (3 months). In November, the entire 'types of biases' section was moved to the article on bias unchanged, the edit being done in order to brevify policy, rather than in objection to any part of it. (descriptions removed from policy, which is wikilinked to 'bias' and added to bias verbatim)

So on the basis that the wording I proposed was accepted (except for one change of word) and remained in policy by community consensus, until the entire section was cutpasted (still intact) to a linked article, I'm going to stick with the description in that diff, as a good policy-based explanation.

What follows is therefore a further explanation of the difference, and a more full description of the issue and our handling.

Explanation of NPOV vs. SPOV - NPOV/SPOV is a conflict that can arise when scientific views on a matter exist, but are not the sole significant view.

In describing matters, we give each view on the matter its due prominence, or weight ("WP:WEIGHT"). When a view is 'mainstream', - ie, widespread, highly accepted by reliable sources, propounded by authoritative voices that themselves are well respected, and non-controversial (there is little authoritative dissent in reliable sources) - then this means it will usually have a significant prominence, or even at times be the sole view stated.

Scientific views often have this standing on subjects they touch upon.

Therefore when articles touch on scientific matters, the most usual result is that the scientific view gains a high prominence where relevant in the description. In many cases this may end up being the sole significant view, in which case it is the sole view described. (eg, the sole significant view why rocks fall down that I am aware of, is in fact the scientific view - gravity. The sole significant view why people catch 'flu is the scientific view - micro-organisms)

The critical factor is, we do not in fact give science this prominence because "science says so" (SPOV), even though it may seem that way. We do it because the scientific view is the sole, or supermajority view on the matter in this instance (NPOV). In other words, although it may look as if we do it because "science says so", in fact we are doing it because it is also a neutral encyclopedic representation of the significant views, for this specific subject.

NPOV is mandatory in all cases. So although the two often agree, NPOV does in fact takes priority (in all cases on Misplaced Pages articles) when there is doubt. For this reason, when there is more than one significant view, we do not just state one view only, or disparage (or flatter) the alternative. Each is neutrally described and given its due weight, as is the balance and mutual perspectives between them.

Practical implications - In practical terms, the main implication of this is that the extent to which the scientific view is given due weight in articles, does not extend so far that there is an actual failure to give due weight or fair mention to significant other views (when they exist), or imply we therefore represent them non-neutrally. If this happens, then there is a NPOV problem, and we need to address it.

No matter how widely respected and correct the scientific view may be, or how much one does not consider some other view or notion to have merit, NPOV is still non-negotiable and mandatory (highly quoted words), and we still give fair and appropriate coverage with due weight to the various significant (ie non-tiny) views. The article needs to describe each significant (non-tiny-minority) view with the following at least in mind:

• Neutrally - neither too much nor too little emphasis and mention.
• As views worthy of being documented and described in their own right as opposed to "through someone elses eyes". (This signifies how we present a topic or viewpoint in the article: even the article on homeopathy must leave the reader able to understand how homeopathists see it, as well as how and why scientists disagree. Likewise, descriptions of the topic by notable proponents, critics, or commentators, or significant other views, are each given neutral balanced presentation also. We present each view in a way that correctly represents the view or words of its speaker, so to speak. We seek to attribute the words of notable proponents or critics to those people, rather than imply they are "the truth" or "the voice of Misplaced Pages", so that at the end of the article the reader has a good neutral grasp of the topic, not just of one viewpoint's description of the topic.)
• Without disparagement or flattery, whether visible or subtle
• Without selective representation or citing to make it look deliberately worse than is fair (eg, by choosing weak positive cites, or strong negative ones, or juxtaposing them in a way that deliberately makes the viewpoint seem more flimsy), or deliberately better than is fair (by doing the reverse).

NPOV covers both what is said, and how it is said, since both are important.

The sections "fairness of tone" and "let the facts speak for themselves" cover this more fully.

Last, and briefly, it's worth noting the arbcom case on pseudoscience contains useful clarifications, covering experts and subjects requiring expertize, different types of "alternative view", and other topics related to the use of scientific and other views within NPOV. The main principle stated by arbcom in interpreting NPOV was again similar to the above:

"NPOV requires fair representation of significant alternatives to scientific orthodoxy. Significant alternatives, in this case, refers to legitimate scientific disagreement, as opposed to pseudoscience."

And thanks for the question!

Glad to answer :)

FT2  00:05, 3 December 2007 (UTC)

No.

FT2  05:56, 2 December 2007 (UTC)

Short version - I'm going to say right up front it is a point of ethics and integrity to not ever edit policy in the manner you describe. If someone does this in bad faith, it would be looked on as a serious attempt to game the system, and serious evidence that they knew their conduct was wrong or against norms, but attempted to anyhow. And the community will probably have a lot less tolerance for failure to quickly get the point.

That said, the nature of policy makes the question rather more interesting than it might seem.... and merits a rather more considered answer. The ultimate deciding factors (to my mind) are not the most immediate ones.

Policies and guidelines - Policies and guidelines are intended to document consensus and communal norms; they do not create it. In the classic description, when we find a norm that is widely respected and followed, we document it so others know about it and don't slip up. This is what is meant when editors talk about the spirit of policy (ie, what the intention of the community was to achieve), their role as principles, the existance of IAR and WP:LAWYER, and so on.

So there's a real irony in that a user editing a policy page has actually achieved nothing. They cannot in fact "change" policy by doing so, because all policy is, is a codification of a communal norm, and it is the communal norm they are judged against; the codification is merely a shortcut to describe that. What they have done is shown bad faith, and disrupted the process for others. Over time some core policy pages have been watered down by those with vested interests in reducing their clarity (WP:SOCK is perhaps a prime target for this); even so the norm has visibly remained unaffected by this untoward influence.

On the other hand my own major refactor and occasional significant work on WP:NPOV - perhaps our most highly valued policy - which were all done as WP:BRD edits, gained approval and 18 months on, are still the stable version. Many policy page edits by many users are valuable, or productive (those that aren't are often quickly reverted). So also, some policy page edits are disruptive, some are not. One has to consider this carefully when considering the pros and cons of different approaches to their editing.

Own handling of the situation described - In my own case, I have actually hit the problem you describe. A quick story:

Policy edits are often suggested by disputes and other events that show the policy is not sufficiently accurate, or helpful, to users. (See for example my answer to #(17) changes to policies, rules, and practices for more on this.)

In July 2007 I hit a problem. I was handling the serious warrior and gamer user:DPeterson, who was at the time at Arbcom (now banned). I was presenting evidence of a year long campaign of virulent gaming and sock warring by the editor. At the time, the guideline Misplaced Pages:Gaming the system was a short section included in WP:POINT , and as such it only covered the couple of aspects of gaming that counted as "disruption to prove a point". To me it was obvious that had there been a full guideline on gaming, the good-faith editors in this war might have seen it, recognized the patterns and found it much easier to formulate their problem at ANI months earlier. So I wrote a full guideline and that was fine.

The problem was, at the same time I was also watching a second situation and it seemed likely I might have to take another edit warrior to arbitration (mediation having been rejected)... and I had drawn in part from that user's conduct to pick the examples added to WP:GAME. So I faced the exact problem described: how would it look if I took an editor to Arbcom, for gaming, having a week beforehand written a guideline specifically identifying his actions as gaming :) Disclosing this publicly would cause conflict; I was not yet certain a case would be needed though.

To ensure no bad faith could be inferred, and that if noticed the user could be reassured it was not intended to be used later against him, I took the extreme measure of writing the following note into the edit of the 'gaming' guideline itself:

... <!-- NOTE: I HAVE A PENDING DISPUTE RESOLUTION CASE WHERE THIS IS AN ISSUE. THE ABOVE EXAMPLE IS TAKEN FROM THAT CASE. THIS NOTE IS ADDED TO STATE THAT I WILL NOT BE PLACING RELIANCE UPON MY OWN EDITING OF A PROJECT PAGE, IN ANY DISPUTE RESOLUTION ARISING, AS "EVIDENCE" IN THAT CASE. THIS NOTE CAN BE REMOVED IN AUGUST 2007. NOTE ALSO COPIED TO TALK PAGE FOR VISIBILITY. - FT2 - --> ...

(I think I might've actually forgotten to put a copy on the talk page, which is slightly embarrassing.)

When the DPeterson case looked like being a highly complex one, I went back and modified this note, to read "September 2007" rather than "August 2007" as the removal date instead .)

So in brief, I'm indeed familiar with the scenario, both from having been there, and also from having seen it play out - in both good and bad faith.

Personal view, should project pages be open to editing? - The community has a strong view that it might be good, and also strong views that it makes policy "akin to legislation". It has not happened yet. My personal view is, consensus will decide. But I do have a stance.

There is great structural advantage to a project page being able to be edited. As noted above, it allows fixes, adaptations and enhancements that might otherwise never happen due to bureaucracy. It keeps us a bit more fluid, a bit less prone to stagnancy. Whilst bad edits are routinely fixed, however, borderline ones may not be, and policies are so central as to be relied on by everybody and a prime target for agenda-driven editing. A questionable but borderline edit can linger some days in various cases - enough time to misinform many readers. And yet, editability is extremely desirable and so is above average stability.

Overall we cope. It's quirky to be that fluid on your main policies, but in fact little harm happens, due to the number of eyeballs and high buy-in they have. We may communally find in future that more formality and less fluidity serves us well, but (see WP:BUREAU, WP:CREEP) at this time the community has not yet shown a willingness to make that decision.

So given a flat choice, I'd keep as we are -- unless we eventually gain an ability to set finer graduations of control over pages, or flagged revisions arrive for example, which would allow more options. But I'd prefer "open to all" than "locked to all but sysops" if I had a choice. Our ethos of 'equal and open respect' matters more, and if the price is more vigilance and some edits to fix... realistically almost no actual harm is usually done in practice. It's an annoyance, and the positive signs it sends and the benefits we gain, may merit standing some small annoyance in order to preserve that ethos that "anyone can edit" even these, of all pages.

Caveat - (And this almost invariably means we will also be likely to tighten up on problem conduct on live policy and guideline pages more than many others, since these impact more on the project than many others. A user who edit wars via 3RR on WP:OR to get "Their version" in, which is not the community's view, will probably be seen rather more seriously and get fewer chances, than if they had done it on some other page, most likely. But either way the community is more than able to handle it.)

This proposed finding relates to the Durova case, in which a number of users had a mailing list on which they did 'sleuthing'. Sleuthing in this context appears to have been discussion of who might be a sock-puppet of whom, based on behaviors, coincidences and other editing patterns allegedly spotted by its members. Whilst it may have had effective results in some cases (for all I know), the present case suggests that the self-declared sleuths were too willing to assume, too simplistic and self-sure, and easily swept along by their own views without thought for cold outside checking by uninvolved level headed reviewers, or that the community might have wiser voices to listen to their findings. As a result a long term user !! (talk · contribs · deleted contribs · logs · filter log · block user · block log) was briefly blocked on what arbcom described as a serious lack of good evidence.

Arbcom case summaries aren't law, but they will be accorded a certain respect, quoted if useful, and looked at by users. An item like this would be in the principles section - a finding that is a generally usable and useful principle that can be relied upon. So a finding like this needs careful construction and consideration.

Unfortunately there are a number of serious problems with this proposal, and I would oppose it. Most were evident on a moment's reading:

1. The use of the word "protected" is very ambiguous. It is unclear how exactly meta activities would be "protected" and what from, for example. It's unclear what the word means in this sentence: "Voting in an arbcom election is 'protected' by registering"? What other "meta-editing activities" are being "protected" by registering? This would be unhelpfully unclear to many users. If it has a useful meaning, it needs a rewrite to express it.
    
2. It's conceptually mistaken. Registering in fact offers no "protection" against flawed allegations of this kind. Registering does not appear to have "protected" user:!! in this case, nor is it clear how registering could have been the slightest help in protecting him. In fact, since registering hides information that could be used by any user to test allegations, the opposite may be slightly true for good-faith users.
    
3. It's a poor response to the case in hand. As it happens, user:!! wasn't trying to exercise any "meta editing activities" (another inadequately defined term). His contributions show he was working on Featured Article review, DYK, cricket related articles and biographies in the days previous to the event.
    
4. It's unclear as to subject/object. The proposal ("meta-editing activities are best protected by registering than by sleuthing") is capable of reasonably sustaining at least three different interpretations by users. (We have seen the problems that can happen due to disregarding the possibility of incorrect interpretation in other cases.) Is the subject (the "thing being protected by registering") in this wording, the "meta-editing activity", the user, or the user's right to engage in the activity? This needs to be better defined. In other words, is it saying that:

   1. Misplaced Pages meta-editing activities such as arbcom elections (etc) will be protected if people register,
   2. Users who wish to engage in Misplaced Pages meta-editing activities will be best protected if they register, or
   3. Users can best protect their right to engage in meta-editing activities if they register?
       

5. I'm also hesitant to endorse a finding that might eat away at a core Misplaced Pages community principle that anyone can edit, even without registering. It's not clear how editing anonymously would make someone more of a target for "sleuthing" or in need of more "protection", and if they aren't at more risk somehow, we shouldn't be implying they are. Arbcom has ruled repeatedly that "All editors deserve to be treated with the utmost of respect by administrators" and blocking should be used with utmost care only. Registering is desirable for many reasons, but administrators are expected to act with thought: "protection" from unthought-out and unjustified blocks shouldn't be one of them.
   
   Comment: A significant contribution in fact comes from unregistered users (see below). A proposal that implies we allow them to edit unregistered but if they don't register they can expect to have more problems or be considered to be 'less valid users' somehow, goes against many long standing Misplaced Pages communal traditions. We exclude IP users from some activities, but we welcome their editing and other involvement completely.

   • (The source for this is a 2007 academic study of Wiki editing, which examining anonymous editing. It noted that in fact, anonymous editors were some of Misplaced Pages's most productive adders of valid content, and stated that, "Surprisingly ... we find the highest quality from the vast numbers of anonymous 'Good Samaritans' who contribute only once." )
      

6. Arbcom tend to avoid unclear terms. I'd possibly look for a better term than "sleuthing" that describes more precisely the activity being addressed.
    
7. It is logically unsound. The essence of this 'sleuthing' was the piecing together of behavioral clues in users edits, to conclude some user was a sockpuppet, either of an unknown user, or some other person. It's not at all clear how the question "are they registered" would impact on (or inter-relate with) that kind of behavioral examination and digging.

   • In more detail, unless some bad faith is assumed such as "sock puppeteers are more likely to use IPs than accounts; registering will tend to protect one against allegations", it's not clear how the two might relate. And in fact experienced sock users are more likely to use accounts, not less, so even that it doesn't make sense.

While this proposal doesn't work for me, if this question means you have a view on what we should be doing on this topic, that hasn't been noted yet, feel free to let me know and I'll take a further look, either here or on my talk page as relevant.

I've merged and covered two issues here, the question above, and also, the further question "arbcom recusals and access to discussion". I'll probably de-merge these in a while.

One line answer -- 1/ Clear and strong oppose on principle, to ex-Arbcom members keeping access to the list (or any other arbcom discussion methods) after their term is over. 2/ Clear and strong oppose to non-Arbcom (non-voted) users being privy to Arbcom discussions other than with exceptional and good reason connected with the benefit of the project (eg Foundation lawyer?). 3/ Clear and strong oppose to parties being able to influence, shoulder-read, or be in the frame, in cases where they have involvement, but concerns over the best way to achieve that technically in practice. Arbcom is not a social club. Despite the experience ex members may be able to provide, list members should be the appointed and present-day Arbcom only. There may be a very few that the Foundation determines have exceptionally some genuine, serious, good reason to need to be aware of Arbcom email traffic and general discussion (eg, Foundation lawyer?). But in general, when a user no longer has genuine cause to be on the mailing list, prompt removal without exception is appropriate (and would be sought in my case even if not generally the norm for others, if appointed), as well as removal of access to any other restricted discussions acquired due to the role.

More... - There are good reasons which might be put forward, why allowing such access to the arbcom mailing list might be valid. Some arguments include:

• Past Arbcom members are trusted and have been trusted to see such cases anyway for years
• Past Arbcom members can advise and help
• There are non-voted users who have good reason, and non-voted users are rarely appointed without good reason. (Mike Godwin, the Foundation's lawyer, might be one example of a non-voted person who might be argued to need sight of a mailing list that may have to handle controversial conduct and oversight matters. Jimmy Wales himself is another)

I can see the merit that a small and restricted number of non-arbitrators might need access, for legal and other reasons. But we should be open who they are, and the nature of their involvement, it should be kept small, and their stance is at most, input on their specific areas of interest. Thus (for example) I'd endorse a legal overview, because I can see the case for it - arbcom deals with oversight and conduct matters some of which may be legally serious. They may not realize a matter is legally significant, so leaving it to them to seek input might result in a serious matter being misjudged through ignorance. But if allowed access, that person is there for legal overview as needed, and to judge a user's conduct or proposal in legal terms, for example..... they are not there to be an auxilliary arbitrator.

But overall no. A clear and strong oppose. When my term is up (if appointed) I expect to be removed from all privileged access I may have been given as part of (and in order to do) my role. If Arbcom want to consult me, or ask my opinion in any matter, that's always welcome. But it should not be presumed, nor should ex-arbitrators be a dead weight on the shoulders of new arbcom members with their new insights, having to handle the shades of Arbcom past members too. A classic case of travelling light. They can always be asked for discussion/insight on a specific matter or principle any time, if ever useful, like any other editor. They don't need access to the list for that.

Likewise the committee may decide to give some member Oversight or Checkuser access, to view material for Arbcom cases. But when given it should be explicitly stated specified if this access is for the role, and terminates when the role does, or if it is intended to endure, so there is no doubt 'up front'. (Although there is less concern in this case, since the presumption is trust will persist and it is a tool, not a line into case discussions, which is being given.)

(Late added clarification: One issue worth considering, is, ex-arbitrators as a check and balance. I thought about this when first answering, decided it didn't swing the decision, and left it at that. I've decided to fill that in somewhat. Arbcom is elected in 3 tranches over a 3 year period. In any given year, only a third of Arbcom is replaced. Of those, sometimes there may be a re-election of an existing arbitrator. So we really don't replace much of Arbcom each year. The 2008 Arbcom will be 2/3 made up of users who have been there a year or more. The experience doesn't suddenly leave Arbcom in January. It's more likely "business much as usual, with most of the committee intact, a minority leave only". Given that, I stand by my view above. Arbcom is best to thank and respect its ex members - and let them go when their tour of duty is up. The 15 to 18 on the Committee and their experience and qualities - including experience when to seek advice - is (or should well be) its own strong set of checks and balances. Staggered membership and arb experience seriously strengthens that. And again, there's no good need, no essential benefit, travelling light is easier work. Arbcom is a job and focus. We will have an arbcom of 15 to 18 or so including many with long experience - that's a good number for the job. More than enough for good insights and safeguards to arise. 30 to 40 (as it seems it may be in a year or so if ex-members remain) isn't. Not that this trumps the previous arguments which are the main ones for "oppose". It just addresses a "reason to keep" that might be considered, and dismissed as insufficient.)

Arbitrators involved in cases having read access or access to discussion - The second point you raise has troubled me too, enough to have asked advice already on this (see below for details). There are two cases here - arbitrators as parties whose conduct is being questioned, and arbitrators involved in a case to present evidence or views. Both have problems. I don't have answers, but I can name some of the problems.

In the former case (parties whose conduct is being questioned), would arbcom criticize and accept a case against one of their own? How much should the case relate to their work or their (possibly questionable) conduct before they are removed from access during the case discussion? Who would decide (Arbcom? Community (how)? Jimmy Wales?) And can this be abused by people seeking to retaliate or cause disruption? (For example, bad-faith assertions used to 'demand' recusal based on a spurious claim the arbitrator was involved?)

In the latter case, what if an arbitrator is involved because they are presenting evidence or an outside view, or some non-fault role that none the less might be seen as taking a side (ie main evidence or outside view) in a case? This has troubled me too. Evidence might be perceived to be favored because of the presenter, whether or not it was in fact the quality of evidence that was the reason. Certainly a person who states a non-trivial view is unlikely to seem neutral in relation to both parties. So there's a problem. I don't know if recusal would in fact be strong enough to ensure visible independence, but no better answer has gained communal consensus in the several years arbcom has existed. I'm almost tempted to say they shouldn't receive emails concerning that case, in order that their recusal and isolation from it as arbitrator is complete, but that kind of filtering is probably not technically possible, and general removal would seriously disrupt their involvement in the other 5-10 cases going on simultaneously. And in any case mailing lists are only one of many forms of communication where cases get discussed.

The other problem is the effect of "chill". A reader implies a degree of reduced freedom to speak freely of that person. If the reader is a possible "fault party" then if one is strictly seeking integrity, criticism must be without their seeing it. They are removed, their case is discussed, they are told the result and allowed back after (if appropriate). If they are a no-fault party and they are presenting evidence or a view, then harm is less likely from reading comments on it, and criticism probably will be freely spoken, but it's still a concern. But once they choose to present evidence at the case, then anything arbcom discuss related to assessing their evidence or setting it against others, becomes again, a possible risk of undue effect. Since I don't think non-current arbitrators need list access, the risk is limited to just the current members of the committee, pretty much. But that is no guarantee problems and conficts between users as parties/presenters, and arbitrators, will not exist.

These issues arise in the real world, too, except there in general, 1/ a person chooses a career as a bringer of cases or a hearer of them, and this usually precludes being both, and 2/ there are many courts, whereas here any user may open an arbitration case and wide views and evidence are valued, and we have only one source of Arbitration. We'd need to expand Arbcom or have multiple circuits , to create extra capacity, if we went that route. But that would be one option perhaps, a two-circuit Arbcom allowing more capacity, and also allowing an Arbitrator on circuit #1 to be involved as a party, in a case on circuit #2, without disruption. And even that approach would have problems of its own, too.

I don't know what is possible or best here. Trust and integrity is absolutely essential in this; can these alone be enough? But the visibility of it is also important, and non-arbitrators have no way to tell whether (and how well) these things are being taken care of.

Hildanknight used a good word discussing the Singapore elections yesterday: "Incorruptible". I think that's what's needed. Arbitrators so strict in their self-managing, that even if they were able to access such matters they wouldn't use it, nor would others be influenced by them. That kind of belief in human nature of a small group is a question many will find unacceptable, and it's hard to argue; that's a real concern.

I leave the question open, but if you have ideas that might be a practical solution, I'm seriously interested too. A more personal note is below.

Thanks for an important question. I'll get to the others shortly.

FT2  22:10, 3 December 2007 (UTC)

Personal Note - If appointed, I am already decided on my measures. I will expect removal from any restricted access after my term is over, even if others do not require it, unless the community itself has openly determined otherwise, and I will be exceptionally deliberate not just to recuse, but to avoid all comment or discussion of cases with other Arbcom members, if I am presenting the evidence for a case. An arbcom member cannot and should never be, both case presenter, and also arbitrator/jury. That would be a rather definite no.

In fact it was a situation I took so seriously, I emailed another respected administrator (the person who nominated at my RFA in fact) on November 6 to ask specifically their advice. My email read in part:

"In these cases, I in fairness both be the one documenting and preparing the case statement and evidence, and also an arbitrator hearing the case. On such cases I would obviously recuse and not participate. The question is, as a possible arbcom position, is that sufficient or should I do more? And does that stance need stating more explicitly or do others take it as read that this would be what happens?"

I am sure the administrator concerned will be pleased to verify this; accordingly I have linked to him above in case anyone wants to have an idea who I asked. The response I got was to recuse.

• Note: I verify that FT2 contacted me on this regard. ≈ jossi ≈ (talk) 18:24, 3 December 2007 (UTC)